diff options
Diffstat (limited to 'lib/erl_interface')
50 files changed, 2113 insertions, 907 deletions
diff --git a/lib/erl_interface/aclocal.m4 b/lib/erl_interface/aclocal.m4 index 9578cd35c4..25f40944e7 100644 --- a/lib/erl_interface/aclocal.m4 +++ b/lib/erl_interface/aclocal.m4 @@ -1,7 +1,7 @@ dnl dnl %CopyrightBegin% dnl -dnl Copyright Ericsson AB 1998-2012. All Rights Reserved. +dnl Copyright Ericsson AB 1998-2013. All Rights Reserved. dnl dnl The contents of this file are subject to the Erlang Public License, dnl Version 1.1, (the "License"); you may not use this file except in @@ -1849,6 +1849,31 @@ case $erl_gethrvtime in esac ])dnl +dnl ---------------------------------------------------------------------- +dnl +dnl LM_TRY_ENABLE_CFLAG +dnl +dnl +dnl Tries a CFLAG and sees if it can be enabled without compiler errors +dnl $1: textual cflag to add +dnl $2: variable to store the modified CFLAG in +dnl Usage example LM_TRY_ENABLE_CFLAG([-Werror=return-type], [CFLAGS]) +dnl +dnl +AC_DEFUN([LM_TRY_ENABLE_CFLAG], [ + AC_MSG_CHECKING([if we can add $1 to $2 (via CFLAGS)]) + saved_CFLAGS=$CFLAGS; + CFLAGS="$1 $$2"; + AC_TRY_COMPILE([],[return 0;],can_enable_flag=true,can_enable_flag=false) + CFLAGS=$saved_CFLAGS; + if test "X$can_enable_flag" = "Xtrue"; then + AC_MSG_RESULT([yes]) + AS_VAR_SET($2, "$1 $$2") + else + AC_MSG_RESULT([no]) + fi +]) + dnl ERL_TRY_LINK_JAVA(CLASSES, FUNCTION-BODY dnl [ACTION_IF_FOUND [, ACTION-IF-NOT-FOUND]]) dnl Freely inspired by AC_TRY_LINK. (Maybe better to create a diff --git a/lib/erl_interface/configure.in b/lib/erl_interface/configure.in index 97f1cff345..d511f2e240 100644 --- a/lib/erl_interface/configure.in +++ b/lib/erl_interface/configure.in @@ -273,6 +273,8 @@ esac AC_SUBST(WFLAGS) if test "x$GCC" = xyes; then + # Treat certain GCC warnings as errors + LM_TRY_ENABLE_CFLAG([-Werror=return-type], [WERRORFLAGS]) WFLAGS="-Wall -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wnested-externs -Winline" # check which GCC version GCC_VERSION=`$CC -v 2>&1 | sed -n 's/gcc version //p'` @@ -287,8 +289,10 @@ if test "x$GCC" = xyes; then *) WFLAGS="$WFLAGS -fno-strict-aliasing";; esac + CFLAGS="$WERRORFLAGS $CFLAGS" else WFLAGS="" + WERRORFLAGS="" fi # --------------------------------------------------------------------------- diff --git a/lib/erl_interface/doc/src/ei.xml b/lib/erl_interface/doc/src/ei.xml index 539e16d837..dfe181bd1d 100644 --- a/lib/erl_interface/doc/src/ei.xml +++ b/lib/erl_interface/doc/src/ei.xml @@ -4,7 +4,7 @@ <cref> <header> <copyright> - <year>2001</year><year>2011</year> + <year>2001</year><year>2013</year> <holder>Ericsson AB. All Rights Reserved.</holder> </copyright> <legalnotice> @@ -82,6 +82,28 @@ function returns the size required (note that for strings an extra byte is needed for the 0 string terminator).</p> </description> + <section> + <title>DATA TYPES</title> + + <taglist> + <tag><marker id="erlang_char_encoding"/>erlang_char_encoding</tag> + <item> + <p/> + <code type="none"> +typedef enum { + ERLANG_ASCII = 1, + ERLANG_LATIN1 = 2, + ERLANG_UTF8 = 4 +}erlang_char_encoding; +</code> + <p>The character encodings used for atoms. <c>ERLANG_ASCII</c> represents 7-bit ASCII. + Latin1 and UTF8 are different extensions of 7-bit ASCII. All 7-bit ASCII characters + are valid Latin1 and UTF8 characters. ASCII and Latin1 both represent each character + by one byte. A UTF8 character can consist of one to four bytes. Note that these + constants are bit-flags and can be combined with bitwise-or.</p> + </item> + </taglist> + </section> <funcs> <func> <name><ret>void</ret><nametext>ei_set_compat_rel(release_number)</nametext></name> @@ -225,12 +247,32 @@ <fsummary>Encode an atom</fsummary> <desc> <p>Encodes an atom in the binary format. The <c><![CDATA[p]]></c> parameter - is the name of the atom. Only upto <c><![CDATA[MAXATOMLEN]]></c> bytes + is the name of the atom in latin1 encoding. Only upto <c>MAXATOMLEN-1</c> bytes are encoded. The name should be zero-terminated, except for the <c><![CDATA[ei_x_encode_atom_len()]]></c> function.</p> </desc> </func> <func> + <name><ret>int</ret><nametext>ei_encode_atom_as(char *buf, int *index, const char *p, erlang_char_encoding from_enc, erlang_char_encoding to_enc)</nametext></name> + <name><ret>int</ret><nametext>ei_encode_atom_len_as(char *buf, int *index, const char *p, int len, erlang_char_encoding from_enc, erlang_char_encoding to_enc)</nametext></name> + <name><ret>int</ret><nametext>ei_x_encode_atom_as(ei_x_buff* x, const char *p, erlang_char_encoding from_enc, erlang_char_encoding to_enc)</nametext></name> + <name><ret>int</ret><nametext>ei_x_encode_atom_len_as(ei_x_buff* x, const char *p, int len, erlang_char_encoding from_enc, erlang_char_encoding to_enc)</nametext></name> + <fsummary>Encode an atom</fsummary> + <desc> + <p>Encodes an atom in the binary format with character encoding + <c><seealso marker="#erlang_char_encoding">to_enc</seealso></c> (latin1 or utf8). + The <c>p</c> parameter is the name of the atom with character encoding + <c><seealso marker="#erlang_char_encoding">from_enc</seealso></c> (ascii, latin1 or utf8). + The name must either be zero-terminated or a function variant with a <c>len</c> + parameter must be used.</p> + <p>The encoding will fail if <c>p</c> is not a valid string in encoding <c>from_enc</c>, + if the string is too long or if it can not be represented with character encoding <c>to_enc</c>.</p> + <p>These functions were introduced in R16 release of Erlang/OTP as part of a first step + to support UTF8 atoms. Atoms encoded with <c>ERLANG_UTF8</c> + can not be decoded by earlier releases than R16.</p> + </desc> + </func> + <func> <name><ret>int</ret><nametext>ei_encode_binary(char *buf, int *index, const void *p, long len)</nametext></name> <name><ret>int</ret><nametext>ei_x_encode_binary(ei_x_buff* x, const void *p, long len)</nametext></name> <fsummary>Encode a binary</fsummary> @@ -490,11 +532,32 @@ ei_x_encode_empty_list(&x); <fsummary>Decode an atom</fsummary> <desc> <p>This function decodes an atom from the binary format. The - name of the atom is placed at <c><![CDATA[p]]></c>. There can be at most + null terminated name of the atom is placed at <c><![CDATA[p]]></c>. There can be at most <c><![CDATA[MAXATOMLEN]]></c> bytes placed in the buffer.</p> </desc> </func> <func> + <name><ret>int</ret><nametext>ei_decode_atom_as(const char *buf, int *index, char *p, int plen, erlang_char_encoding want, erlang_char_encoding* was, erlang_char_encoding* result)</nametext></name> + <fsummary>Decode an atom</fsummary> + <desc> + <p>This function decodes an atom from the binary format. The + null terminated name of the atom is placed in buffer at <c>p</c> of length + <c>plen</c> bytes.</p> + <p>The wanted string encoding is specified by <c><seealso marker="#erlang_char_encoding"> + want</seealso></c>. The original encoding used in the + binary format (latin1 or utf8) can be obtained from <c>*was</c>. The actual encoding of the resulting string + (7-bit ascii, latin1 or utf8) can be obtained from <c>*result</c>. Both <c>was</c> and <c>result</c> can be <c>NULL</c>. + + <c>*result</c> may differ from <c>want</c> if <c>want</c> is a bitwise-or'd combination like + <c>ERLANG_LATIN1|ERLANG_UTF8</c> or if <c>*result</c> turn out to be pure 7-bit ascii + (compatible with both latin1 and utf8).</p> + <p>This function fails if the atom is too long for the buffer + or if it can not be represented with encoding <c>want</c>.</p> + <p>This function was introduced in R16 release of Erlang/OTP as part of a first step + to support UTF8 atoms.</p> + </desc> + </func> + <func> <name><ret>int</ret><nametext>ei_decode_binary(const char *buf, int *index, void *p, long *len)</nametext></name> <fsummary>Decode a binary</fsummary> <desc> diff --git a/lib/erl_interface/doc/src/erl_eterm.xml b/lib/erl_interface/doc/src/erl_eterm.xml index f403618c59..a7921ea9d2 100644 --- a/lib/erl_interface/doc/src/erl_eterm.xml +++ b/lib/erl_interface/doc/src/erl_eterm.xml @@ -4,7 +4,7 @@ <cref> <header> <copyright> - <year>1996</year><year>2009</year> + <year>1996</year><year>2013</year> <holder>Ericsson AB. All Rights Reserved.</holder> </copyright> <legalnotice> @@ -77,10 +77,12 @@ </p> <taglist> <tag><c><![CDATA[char *ERL_ATOM_PTR(t)]]></c></tag> + <tag><c><![CDATA[char *ERL_ATOM_PTR_UTF8(t)]]></c></tag> <item>A string representing atom <c><![CDATA[t]]></c>. </item> <tag><c><![CDATA[int ERL_ATOM_SIZE(t)]]></c></tag> - <item>The length (in characters) of atom t.</item> + <tag><c><![CDATA[int ERL_ATOM_SIZE_UTF8(t)]]></c></tag> + <item>The length (in bytes) of atom t.</item> <tag><c><![CDATA[void *ERL_BIN_PTR(t)]]></c></tag> <item>A pointer to the contents of <c><![CDATA[t]]></c></item> <tag><c><![CDATA[int ERL_BIN_SIZE(t)]]></c></tag> @@ -92,6 +94,7 @@ <tag><c><![CDATA[double ERL_FLOAT_VALUE(t)]]></c></tag> <item>The floating point value of <c><![CDATA[t]]></c>.</item> <tag><c><![CDATA[ETERM *ERL_PID_NODE(t)]]></c></tag> + <tag><c><![CDATA[ETERM *ERL_PID_NODE_UTF8(t)]]></c></tag> <item>The Node in pid <c><![CDATA[t]]></c>.</item> <tag><c><![CDATA[int ERL_PID_NUMBER(t)]]></c></tag> <item>The sequence number in pid <c><![CDATA[t]]></c>.</item> @@ -104,6 +107,7 @@ <tag><c><![CDATA[int ERL_PORT_CREATION(t)]]></c></tag> <item>The creation number in port <c><![CDATA[t]]></c>.</item> <tag><c><![CDATA[ETERM *ERL_PORT_NODE(t)]]></c></tag> + <tag><c><![CDATA[ETERM *ERL_PORT_NODE_UTF8(t)]]></c></tag> <item>The node in port <c><![CDATA[t]]></c>.</item> <tag><c><![CDATA[int ERL_REF_NUMBER(t)]]></c></tag> <item>The first part of the reference number in ref <c><![CDATA[t]]></c>. Use @@ -296,7 +300,7 @@ iohead ::= Binary <name><ret>ETERM *</ret><nametext>erl_mk_atom(string)</nametext></name> <fsummary>Creates an atom</fsummary> <type> - <v>char *string;</v> + <v>const char *string;</v> </type> <desc> <p>Creates an atom.</p> @@ -305,10 +309,12 @@ iohead ::= Binary <p>Returns an Erlang term containing an atom. Note that it is the callers responsibility to make sure that <c><![CDATA[string]]></c> contains a valid name for an atom.</p> - <p><c><![CDATA[ERL_ATOM_PTR(atom)]]></c> can be used to retrieve the - atom name (as a string). Note that the string is not - 0-terminated in the atom. <c><![CDATA[ERL_ATOM_SIZE(atom)]]></c>returns - the length of the atom name.</p> + <p><c><![CDATA[ERL_ATOM_PTR(atom)]]></c> and <c><![CDATA[ERL_ATOM_PTR_UTF8(atom)]]></c> + can be used to retrieve the atom name (as a null terminated string). <c><![CDATA[ERL_ATOM_SIZE(atom)]]></c> + and <c><![CDATA[ERL_ATOM_SIZE_UTF8(atom)]]></c> returns the length of the atom name.</p> + <note><p>Note that the UTF8 variants were introduced in Erlang/OTP releases R16 + and the string returned by <c>ERL_ATOM_PTR(atom)</c> was not null terminated on older releases.</p> + </note> </desc> </func> <func> diff --git a/lib/erl_interface/doc/src/notes.xml b/lib/erl_interface/doc/src/notes.xml index f0a9b336ff..7cf3d46d88 100644 --- a/lib/erl_interface/doc/src/notes.xml +++ b/lib/erl_interface/doc/src/notes.xml @@ -30,6 +30,33 @@ </header> <p>This document describes the changes made to the Erl_interface application.</p> +<section><title>Erl_Interface 3.7.10</title> + + <section><title>Improvements and New Features</title> + <list> + <item> + <p>Where necessary a comment stating encoding has been + added to Erlang files. The comment is meant to be removed + in Erlang/OTP R17B when UTF-8 becomes the default + encoding. </p> + <p> + Own Id: OTP-10630</p> + </item> + <item> + <p> + Limited support for unicode atoms in the external format + and in the internal representation of the vm. This is a + preparative feature in order to support communication + with future releases of Erlang/OTP that may create + unicode atoms.</p> + <p> + Own Id: OTP-10753</p> + </item> + </list> + </section> + +</section> + <section><title>Erl_Interface 3.7.9</title> <section><title>Improvements and New Features</title> diff --git a/lib/erl_interface/include/ei.h b/lib/erl_interface/include/ei.h index ae815b414a..66dc64a69d 100644 --- a/lib/erl_interface/include/ei.h +++ b/lib/erl_interface/include/ei.h @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2010. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -115,6 +115,9 @@ #define ERL_FLOAT_EXT 'c' #define NEW_FLOAT_EXT 'F' #define ERL_ATOM_EXT 'd' +#define ERL_SMALL_ATOM_EXT 's' +#define ERL_ATOM_UTF8_EXT 'v' +#define ERL_SMALL_ATOM_UTF8_EXT 'w' #define ERL_REFERENCE_EXT 'e' #define ERL_NEW_REFERENCE_EXT 'r' #define ERL_PORT_EXT 'f' @@ -183,12 +186,20 @@ extern volatile int __erl_errno; #define EI_MAXHOSTNAMELEN 64 #define EI_MAXALIVELEN 63 #define EI_MAX_COOKIE_SIZE 512 -#define MAXATOMLEN 255 +#define MAXATOMLEN (255 + 1) +#define MAXATOMLEN_UTF8 (255*4 + 1) #define MAXNODELEN EI_MAXALIVELEN+1+EI_MAXHOSTNAMELEN +typedef enum { + ERLANG_ASCII = 1, + ERLANG_LATIN1 = 2, + ERLANG_UTF8 = 4, +}erlang_char_encoding; + /* a pid */ typedef struct { - char node[MAXATOMLEN+1]; + char node[MAXATOMLEN_UTF8]; + erlang_char_encoding node_org_enc; unsigned int num; unsigned int serial; unsigned int creation; @@ -196,14 +207,16 @@ typedef struct { /* a port */ typedef struct { - char node[MAXATOMLEN+1]; + char node[MAXATOMLEN_UTF8]; + erlang_char_encoding node_org_enc; unsigned int id; unsigned int creation; } erlang_port; /* a ref */ typedef struct { - char node[MAXATOMLEN+1]; + char node[MAXATOMLEN_UTF8]; + erlang_char_encoding node_org_enc; int len; unsigned int n[3]; unsigned int creation; @@ -223,15 +236,16 @@ typedef struct { long msgtype; erlang_pid from; erlang_pid to; - char toname[MAXATOMLEN+1]; - char cookie[MAXATOMLEN+1]; + char toname[MAXATOMLEN_UTF8]; + char cookie[MAXATOMLEN_UTF8]; erlang_trace token; } erlang_msg; /* a fun */ typedef struct { long arity; - char module[MAXATOMLEN+1]; + char module[MAXATOMLEN_UTF8]; + erlang_char_encoding module_org_enc; char md5[16]; long index; long old_index; @@ -256,7 +270,7 @@ typedef struct { union { long i_val; double d_val; - char atom_name[MAXATOMLEN+1]; + char atom_name[MAXATOMLEN_UTF8]; erlang_pid pid; erlang_port port; erlang_ref ref; @@ -425,9 +439,17 @@ int ei_encode_string_len(char *buf, int *index, const char *p, int len); int ei_x_encode_string(ei_x_buff* x, const char* s); int ei_x_encode_string_len(ei_x_buff* x, const char* s, int len); int ei_encode_atom(char *buf, int *index, const char *p); +int ei_encode_atom_as(char *buf, int *index, const char *p, + erlang_char_encoding from, erlang_char_encoding to); int ei_encode_atom_len(char *buf, int *index, const char *p, int len); +int ei_encode_atom_len_as(char *buf, int *index, const char *p, int len, + erlang_char_encoding from, erlang_char_encoding to); int ei_x_encode_atom(ei_x_buff* x, const char* s); +int ei_x_encode_atom_as(ei_x_buff* x, const char* s, + erlang_char_encoding from, erlang_char_encoding to); int ei_x_encode_atom_len(ei_x_buff* x, const char* s, int len); +int ei_x_encode_atom_len_as(ei_x_buff* x, const char* s, int len, + erlang_char_encoding from, erlang_char_encoding to); int ei_encode_binary(char *buf, int *index, const void *p, long len); int ei_x_encode_binary(ei_x_buff* x, const void* s, int len); int ei_encode_pid(char *buf, int *index, const erlang_pid *p); @@ -477,6 +499,7 @@ int ei_decode_boolean(const char *buf, int *index, int *p); int ei_decode_char(const char *buf, int *index, char *p); int ei_decode_string(const char *buf, int *index, char *p); int ei_decode_atom(const char *buf, int *index, char *p); +int ei_decode_atom_as(const char *buf, int *index, char *p, int destlen, erlang_char_encoding want, erlang_char_encoding* was, erlang_char_encoding* result); int ei_decode_binary(const char *buf, int *index, void *p, long *len); int ei_decode_fun(const char* buf, int* index, erlang_fun* p); void free_fun(erlang_fun* f); diff --git a/lib/erl_interface/include/erl_interface.h b/lib/erl_interface/include/erl_interface.h index 1c4a94700d..7f44cb83bd 100644 --- a/lib/erl_interface/include/erl_interface.h +++ b/lib/erl_interface/include/erl_interface.h @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2009. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -95,19 +95,24 @@ #define ERL_FLOAT_VALUE(x) ((x)->uval.fval.f) -#define ERL_ATOM_PTR(x) ((x)->uval.aval.a) -#define ERL_ATOM_SIZE(x) ((x)->uval.aval.len) +#define ERL_ATOM_PTR(x) erl_atom_ptr_latin1((Erl_Atom_data*) &(x)->uval.aval.d) +#define ERL_ATOM_PTR_UTF8(x) erl_atom_ptr_utf8((Erl_Atom_data*) &(x)->uval.aval.d) +#define ERL_ATOM_SIZE(x) erl_atom_size_latin1((Erl_Atom_data*) &(x)->uval.aval.d) +#define ERL_ATOM_SIZE_UTF8(x) erl_atom_size_utf8((Erl_Atom_data*) &(x)->uval.aval.d) -#define ERL_PID_NODE(x) ((x)->uval.pidval.node) +#define ERL_PID_NODE(x) erl_atom_ptr_latin1((Erl_Atom_data*) &(x)->uval.pidval.node) +#define ERL_PID_NODE_UTF8(x) erl_atom_ptr_utf8((Erl_Atom_data*) &(x)->uval.pidval.node) #define ERL_PID_NUMBER(x) ((x)->uval.pidval.number) #define ERL_PID_SERIAL(x) ((x)->uval.pidval.serial) #define ERL_PID_CREATION(x) ((x)->uval.pidval.creation) -#define ERL_PORT_NODE(x) ((x)->uval.portval.node) +#define ERL_PORT_NODE(x) erl_atom_ptr_latin1((Erl_Atom_data*) &(x)->uval.portval.node) +#define ERL_PORT_NODE_UTF8(x) erl_atom_ptr_utf8((Erl_Atom_data*) &(x)->uval.portval.node) #define ERL_PORT_NUMBER(x) ((x)->uval.portval.number) #define ERL_PORT_CREATION(x) ((x)->uval.portval.creation) -#define ERL_REF_NODE(x) ((x)->uval.refval.node) +#define ERL_REF_NODE(x) erl_atom_ptr_latin1((Erl_Atom_data*) &(x)->uval.refval.node) +#define ERL_REF_NODE_UTF8(x) erl_atom_ptr_utf8((Erl_Atom_data*) &(x)->uval.refval.node) #define ERL_REF_NUMBER(x) ((x)->uval.refval.n[0]) #define ERL_REF_NUMBERS(x) ((x)->uval.refval.n) #define ERL_REF_LEN(x) ((x)->uval.refval.len) @@ -183,14 +188,26 @@ typedef struct { } Erl_Float; typedef struct { + char *utf8; + int lenU; + char *latin1; + int lenL; +} Erl_Atom_data; + +char* erl_atom_ptr_latin1(Erl_Atom_data*); +char* erl_atom_ptr_utf8(Erl_Atom_data*); +int erl_atom_size_latin1(Erl_Atom_data*); +int erl_atom_size_utf8(Erl_Atom_data*); +char* erl_atom_init_latin1(Erl_Atom_data*, const char*); + +typedef struct { Erl_Header h; - int len; - char *a; + Erl_Atom_data d; } Erl_Atom; typedef struct { Erl_Header h; - char * node; + Erl_Atom_data node; unsigned int number; unsigned int serial; unsigned char creation; @@ -198,14 +215,14 @@ typedef struct { typedef struct { Erl_Header h; - char * node; + Erl_Atom_data node; unsigned int number; unsigned char creation; } Erl_Port; typedef struct { Erl_Header h; - char * node; + Erl_Atom_data node; int len; unsigned int n[3]; unsigned char creation; @@ -289,7 +306,7 @@ typedef struct _eterm { } ETERM; -#define MAXREGLEN 255 /* max length of registered (atom) name */ +#define MAXREGLEN (255*4) /* max length of registered (atom) name */ typedef struct { int type; /* one of the message type constants in eiext.h */ @@ -409,6 +426,7 @@ unsigned char erl_ext_type(unsigned char*); /* Note: returned 'char' before R9C unsigned char *erl_peek_ext(unsigned char*,int); int erl_term_len(ETERM*); +int cmp_latin1_vs_utf8(const char* sL, int lenL, const char* sU, int lenU); /* -------------------------------------------------------------------- */ /* Wrappers around ei functions */ diff --git a/lib/erl_interface/src/Makefile b/lib/erl_interface/src/Makefile index 5f0367bec1..a069aa07ab 100644 --- a/lib/erl_interface/src/Makefile +++ b/lib/erl_interface/src/Makefile @@ -1,7 +1,7 @@ # # %CopyrightBegin% # -# Copyright Ericsson AB 1996-2009. All Rights Reserved. +# Copyright Ericsson AB 1996-2013. All Rights Reserved. # # The contents of this file are subject to the Erlang Public License, # Version 1.1, (the "License"); you may not use this file except in @@ -22,10 +22,11 @@ # FIXME let configure put in this last part TARGET +include $(ERL_TOP)/make/output.mk include $(ERL_TOP)/make/target.mk debug opt shared purify quantify purecov gcov: - $(MAKE) -f $(TARGET)/Makefile TYPE=$@ + $(make_verbose)$(MAKE) -f $(TARGET)/Makefile TYPE=$@ clean depend docs release release_docs tests release_tests check: - $(MAKE) -f $(TARGET)/Makefile $@ + $(make_verbose)$(MAKE) -f $(TARGET)/Makefile $@ diff --git a/lib/erl_interface/src/Makefile.in b/lib/erl_interface/src/Makefile.in index cb41391fe9..ebacc1cee0 100644 --- a/lib/erl_interface/src/Makefile.in +++ b/lib/erl_interface/src/Makefile.in @@ -38,6 +38,8 @@ TARGET = @TARGET@ include ../vsn.mk include $(TARGET)/eidefs.mk +include $(ERL_TOP)/make/output.mk + USING_MINGW=@MIXED_CYGWIN_MINGW@ USING_MSYS_VC==@MIXED_MSYS_VC@ USING_CYGWIN_VC==@MIXED_MSYS_VC@ @@ -98,6 +100,12 @@ LD = @LD@ AR = @AR@ RANLIB = @RANLIB@ +ifeq ($(V),0) +AR_FLAGS=rc +else +AR_FLAGS=rcv +endif + INCFLAGS = -I. -I../include -Iconnect -Iencode -Idecode -Imisc -Iepmd \ -Iregistry -I$(TARGET) @@ -552,16 +560,18 @@ distclean: clean ifeq ($(findstring vxworks,$(TARGET)),vxworks) $(TARGET)/config.h: - echo "/* Generated by Makefile */" > $@ - echo "#define HAVE_STRERROR 1" >> $@ - echo "#define HAVE_SOCKLEN_T 1" >> $@ + $(gen_verbose) + $(V_at)echo "/* Generated by Makefile */" > $@ + $(V_at)echo "#define HAVE_STRERROR 1" >> $@ + $(V_at)echo "#define HAVE_SOCKLEN_T 1" >> $@ endif ifeq ($(findstring ose,$(TARGET)),ose) $(TARGET)/config.h: - echo "/* Generated by Makefile */" > $@ - echo "#define HAVE_STRERROR 1" >> $@ - echo "#define HAVE_SOCKLEN_T 1" >> $@ + $(gen_verbose) + $(V_at)echo "/* Generated by Makefile */" > $@ + $(V_at)echo "#define HAVE_STRERROR 1" >> $@ + $(V_at)echo "#define HAVE_SOCKLEN_T 1" >> $@ endif ########################################################################### @@ -569,19 +579,19 @@ endif ########################################################################### $(ST_OBJDIR)/%.o: %.c - $(CC) $(CFLAGS) -c $< -o $@ + $(V_CC) $(CFLAGS) -c $< -o $@ $(MT_OBJDIR)/%.o: %.c - $(CC) $(MTFLAG) $(CFLAGS) $(THR_DEFS) -c $< -o $@ + $(V_CC) $(MTFLAG) $(CFLAGS) $(THR_DEFS) -c $< -o $@ $(MD_OBJDIR)/%.o: %.c - $(CC) -MD $(CFLAGS) $(THR_DEFS) -c $< -o $@ + $(V_CC) -MD $(CFLAGS) $(THR_DEFS) -c $< -o $@ $(MD_OBJDIR)/%.o: %.c - $(CC) -MD $(CFLAGS) $(THR_DEFS) -c $< -o $@ + $(V_CC) -MD $(CFLAGS) $(THR_DEFS) -c $< -o $@ $(MDD_OBJDIR)/%.o: %.c - $(CC) -MDd $(CFLAGS) $(THR_DEFS) -c $< -o $@ + $(V_CC) -MDd $(CFLAGS) $(THR_DEFS) -c $< -o $@ ########################################################################### # Create directories @@ -598,67 +608,67 @@ ifeq ($(TARGET),win32) # Windows archive creation $(ST_EILIB) : $(ST_EIOBJECTS) - $(AR) -out:$@ $(ST_EIOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(ST_EIOBJECTS) + $(V_RANLIB) $@ $(ST_ERLLIB) : $(ST_ERLOBJECTS) - $(AR) -out:$@ $(ST_ERLOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(ST_ERLOBJECTS) + $(V_RANLIB) $@ $(MT_EILIB) : $(MT_EIOBJECTS) - $(AR) -out:$@ $(MT_EIOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(MT_EIOBJECTS) + $(V_RANLIB) $@ $(MT_ERLLIB) : $(MT_ERLOBJECTS) - $(AR) -out:$@ $(MT_ERLOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(MT_ERLOBJECTS) + $(V_RANLIB) $@ $(MD_EILIB) : $(MD_EIOBJECTS) - $(AR) -out:$@ $(MD_EIOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(MD_EIOBJECTS) + $(V_RANLIB) $@ $(MD_ERLLIB) : $(MD_ERLOBJECTS) - $(AR) -out:$@ $(MD_ERLOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(MD_ERLOBJECTS) + $(V_RANLIB) $@ $(MDD_EILIB) : $(MDD_EIOBJECTS) - $(AR) -out:$@ $(MDD_EIOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(MDD_EIOBJECTS) + $(V_RANLIB) $@ $(MDD_ERLLIB) : $(MDD_ERLOBJECTS) - $(AR) -out:$@ $(MDD_ERLOBJECTS) - $(RANLIB) $@ + $(V_AR) -out:$@ $(MDD_ERLOBJECTS) + $(V_RANLIB) $@ else # Unix archive creation $(ST_EILIB) : $(ST_EIOBJECTS) - rm -f $@ - $(AR) rcv $@ $(ST_EIOBJECTS) + $(V_at)rm -f $@ + $(V_AR) $(AR_FLAGS) $@ $(ST_EIOBJECTS) ifdef RANLIB - $(RANLIB) $@ + $(V_RANLIB) $@ endif $(ST_ERLLIB) : $(ST_ERLOBJECTS) - rm -f $@ - $(AR) rcv $@ $(ST_ERLOBJECTS) + $(V_at)rm -f $@ + $(V_AR) $(AR_FLAGS) $@ $(ST_ERLOBJECTS) ifdef RANLIB - $(RANLIB) $@ + $(V_RANLIB) $@ endif $(MT_EILIB) : $(MT_EIOBJECTS) - rm -f $@ - $(AR) rcv $@ $(MT_EIOBJECTS) + $(V_at)rm -f $@ + $(V_AR) $(AR_FLAGS) $@ $(MT_EIOBJECTS) ifdef RANLIB - $(RANLIB) $@ + $(V_RANLIB) $@ endif $(MT_ERLLIB) : $(MT_ERLOBJECTS) - rm -f $@ - $(AR) rcv $@ $(MT_ERLOBJECTS) + $(V_at)rm -f $@ + $(V_AR) $(AR_FLAGS) $@ $(MT_ERLOBJECTS) ifdef RANLIB - $(RANLIB) $@ + $(V_RANLIB) $@ endif endif @@ -669,18 +679,18 @@ endif ifeq ($(TARGET),win32) $(ERL_CALL): $(ERLCALL) ../include/ei.h $(MD_EILIB) - $(PURIFY) $(CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $(ERLCALL) \ + $(ld_verbose)$(PURIFY) $(CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $(ERLCALL) \ -L$(OBJDIR) -lei_md $(THR_LIBS) $(LIBS) -lsocket else ifeq ($(findstring vxworks,$(TARGET)),vxworks) $(ERL_CALL): $(ST_OBJDIR)/erl_call.o $(ST_OBJDIR)/erl_start.o ../include/ei.h $(ST_EILIB) - $(LD) -r -d -o $@ $(ST_OBJDIR)/erl_call.o $(ST_OBJDIR)/erl_start.o -L$(OBJDIR) -lei $(LIBS) + $(V_LD) -r -d -o $@ $(ST_OBJDIR)/erl_call.o $(ST_OBJDIR)/erl_start.o -L$(OBJDIR) -lei $(LIBS) $(ST_OBJDIR)/erl_call.o: prog/erl_call.c - $(CC) $(CFLAGS) -c $< -o $@ + $(V_CC) $(CFLAGS) -c $< -o $@ $(ST_OBJDIR)/erl_start.o: prog/erl_start.c - $(CC) $(CFLAGS) -c $< -o $@ + $(V_CC) $(CFLAGS) -c $< -o $@ else ifeq ($(findstring ose,$(TARGET)),ose) @@ -688,11 +698,11 @@ $(ERL_CALL): else ifdef THR_DEFS $(ERL_CALL): $(ERLCALL) ../include/ei.h $(MT_EILIB) - $(PURIFY) $(CC) $(PROG_CFLAGS) $(THR_DEFS) $(LDFLAGS) -o $@ $(ERLCALL) \ + $(ld_verbose)$(PURIFY) $(CC) $(PROG_CFLAGS) $(THR_DEFS) $(LDFLAGS) -o $@ $(ERLCALL) \ -L$(OBJDIR) -lei $(THR_LIBS) $(LIBS) else $(ERL_CALL): $(ERLCALL) ../include/ei.h $(ST_EILIB) - $(PURIFY) $(CC) $(PROG_CFLAGS) $(LDFLAGS) -o $@ $(ERLCALL) \ + $(ld_verbose)$(PURIFY) $(CC) $(PROG_CFLAGS) $(LDFLAGS) -o $@ $(ERLCALL) \ -L$(OBJDIR) -lei $(LIBS) endif endif @@ -707,36 +717,36 @@ check: $(FAKE_TARGETS) ifndef THR_DEFS $(ST_OBJDIR)/erl_fake_prog_st$(EXE): prog/erl_fake_prog.c $(ST_ERLLIB) $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lerl_interface -lei \ + $(V_CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lerl_interface -lei \ $(LIBS) $(ST_OBJDIR)/ei_fake_prog_st$(EXE): prog/ei_fake_prog.c $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lei $(LIBS) + $(V_CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lei $(LIBS) $(ST_OBJDIR)/erl_fake_prog_cxx_st$(EXE): prog/erl_fake_prog.c \ $(ST_ERLLIB) $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) \ + $(V_CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) \ -lerl_interface -lei $(LIBS) $(ST_OBJDIR)/ei_fake_prog_cxx_st$(EXE): prog/ei_fake_prog.c $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) -lei $(LIBS) + $(V_CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) -lei $(LIBS) else $(ST_OBJDIR)/erl_fake_prog_st$(EXE): prog/erl_fake_prog.c $(ST_ERLLIB) $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lerl_interface_st -lei_st \ + $(V_CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lerl_interface_st -lei_st \ $(LIBS) $(ST_OBJDIR)/ei_fake_prog_st$(EXE): prog/ei_fake_prog.c $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lei_st $(LIBS) + $(V_CC) $(PROG_CFLAGS) -o $@ $< -L$(OBJDIR) -lei_st $(LIBS) $(ST_OBJDIR)/erl_fake_prog_cxx_st$(EXE): prog/erl_fake_prog.c \ $(ST_ERLLIB) $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) \ + $(V_CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) \ -lerl_interface_st -lei_st $(LIBS) $(ST_OBJDIR)/ei_fake_prog_cxx_st$(EXE): prog/ei_fake_prog.c $(ST_EILIB) - $(CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) -lei_st $(LIBS) + $(V_CC) $(PROG_CFLAGS) -o $@ -xc++ $< -L$(OBJDIR) -lei_st $(LIBS) endif @@ -744,63 +754,63 @@ endif $(MT_OBJDIR)/erl_fake_prog_mt$(EXE): prog/erl_fake_prog.c \ $(MT_ERLLIB) $(MT_EILIB) - $(CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< -L$(OBJDIR) \ + $(V_CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< -L$(OBJDIR) \ -lerl_interface -lei $(THR_LIBS) $(LIBS) $(MT_OBJDIR)/ei_fake_prog_mt$(EXE): prog/ei_fake_prog.c $(MT_EILIB) - $(CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< \ + $(V_CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< \ -L$(OBJDIR) -lei $(THR_LIBS) $(LIBS) $(MT_OBJDIR)/erl_fake_prog_mt_cxx$(EXE): prog/erl_fake_prog.c \ $(MT_ERLLIB) $(MT_EILIB) - $(CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ + $(V_CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ -L$(OBJDIR) -lerl_interface -lei \ $(THR_LIBS) $(LIBS) $(MT_OBJDIR)/ei_fake_prog_mt_cxx$(EXE): prog/ei_fake_prog.c $(MT_EILIB) - $(CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ + $(V_CC) $(MTFLAG) $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ -L$(OBJDIR) -lei $(THR_LIBS) $(LIBS) #### $(MD_OBJDIR)/erl_fake_prog_md$(EXE): prog/erl_fake_prog.c \ $(MD_ERLLIB) $(MD_EILIB) - $(CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< -L$(OBJDIR) \ + $(V_CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< -L$(OBJDIR) \ -lerl_interface_r -lei_r $(THR_LIBS) $(LIBS) $(MD_OBJDIR)/ei_fake_prog_md$(EXE): prog/ei_fake_prog.c $(MD_EILIB) - $(CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< \ + $(V_CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< \ -L$(OBJDIR) -lei_r $(THR_LIBS) $(LIBS) $(MD_OBJDIR)/erl_fake_prog_md_cxx$(EXE): prog/erl_fake_prog.c \ $(MD_ERLLIB) $(MD_EILIB) - $(CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ + $(V_CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ -L$(OBJDIR) -lerl_interface_r -lei_r \ $(THR_LIBS) $(LIBS) $(MD_OBJDIR)/ei_fake_prog_md_cxx$(EXE): prog/ei_fake_prog.c $(MD_EILIB) - $(CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ + $(V_CC) -MD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ -L$(OBJDIR) -lei_r $(THR_LIBS) $(LIBS) #### $(MDD_OBJDIR)/erl_fake_prog_mdd$(EXE): prog/erl_fake_prog.c \ $(MDD_ERLLIB) $(MDD_EILIB) - $(CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< -L$(OBJDIR) \ + $(V_CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< -L$(OBJDIR) \ -lerl_interface_r -lei_r $(THR_LIBS) $(LIBS) $(MDD_OBJDIR)/ei_fake_prog_mdd$(EXE): prog/ei_fake_prog.c $(MDD_EILIB) - $(CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< \ + $(V_CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ $< \ -L$(OBJDIR) -lei_r $(THR_LIBS) $(LIBS) $(MDD_OBJDIR)/erl_fake_prog_mdd_cxx$(EXE): prog/erl_fake_prog.c \ $(MDD_ERLLIB) $(MDD_EILIB) - $(CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ + $(V_CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ -L$(OBJDIR) -lerl_interface_r -lei_r \ $(THR_LIBS) $(LIBS) $(MDD_OBJDIR)/ei_fake_prog_mdd_cxx$(EXE): prog/ei_fake_prog.c $(MDD_EILIB) - $(CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ + $(V_CC) -MDD $(PROG_CFLAGS) $(THR_DEFS) -o $@ -xc++ $< \ -L$(OBJDIR) -lei_r $(THR_LIBS) $(LIBS) ########################################################################### @@ -808,9 +818,10 @@ $(MDD_OBJDIR)/ei_fake_prog_mdd_cxx$(EXE): prog/ei_fake_prog.c $(MDD_EILIB) ########################################################################### depend: - @echo "Generating dependency file depend.mk..." + $(gen_verbose) + $(V_colon)@echo "Generating dependency file depend.mk..." @echo "# Generated dependency rules" > depend.mk; \ - $(CC) $(CFLAGS) -MM $(SOURCES) | \ + $(V_CC) $(CFLAGS) -MM $(SOURCES) | \ sed 's&$(TARGET)&\$$\(TARGET\)&g' | \ sed 's/^.*:/\$$\(ST_OBJDIR\)\/&/' >> depend.mk; \ echo >> depend.mk; \ diff --git a/lib/erl_interface/src/connect/ei_connect.c b/lib/erl_interface/src/connect/ei_connect.c index 34362b4b9f..c1361e169e 100644 --- a/lib/erl_interface/src/connect/ei_connect.c +++ b/lib/erl_interface/src/connect/ei_connect.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2000-2011. All Rights Reserved. + * Copyright Ericsson AB 2000-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -459,6 +459,7 @@ int ei_connect_xinit(ei_cnode* ec, const char *thishostname, /* memmove(&ec->this_ipaddr, thisipaddr, sizeof(ec->this_ipaddr)); */ strcpy(ec->self.node,thisnodename); + ec->self.node_org_enc = ERLANG_LATIN1; ec->self.num = 0; ec->self.serial = 0; ec->self.creation = creation; @@ -1070,7 +1071,7 @@ int ei_rpc(ei_cnode* ec, int fd, char *mod, char *fun, int i, index; ei_term t; erlang_msg msg; - char rex[MAXATOMLEN+1]; + char rex[MAXATOMLEN]; if (ei_rpc_to(ec, fd, mod, fun, inbuf, inbuflen) < 0) { return -1; @@ -1332,7 +1333,9 @@ static int send_name_or_challenge(int fd, char *nodename, | DFLAG_EXTENDED_PIDS_PORTS | DFLAG_FUN_TAGS | DFLAG_NEW_FUN_TAGS - | DFLAG_NEW_FLOATS)); + | DFLAG_NEW_FLOATS + | DFLAG_SMALL_ATOM_TAGS + | DFLAG_UTF8_ATOMS)); if (f_chall) put32be(s, challenge); memcpy(s, nodename, strlen(nodename)); diff --git a/lib/erl_interface/src/connect/ei_connect_int.h b/lib/erl_interface/src/connect/ei_connect_int.h index 3c42b49b82..42ab9b58d7 100644 --- a/lib/erl_interface/src/connect/ei_connect_int.h +++ b/lib/erl_interface/src/connect/ei_connect_int.h @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2010. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -102,6 +102,8 @@ extern int h_errno; #define DFLAG_NEW_FUN_TAGS 0x80 #define DFLAG_EXTENDED_PIDS_PORTS 0x100 #define DFLAG_NEW_FLOATS 0x800 +#define DFLAG_SMALL_ATOM_TAGS 0x4000 +#define DFLAG_UTF8_ATOMS 0x10000 ei_cnode *ei_fd_to_cnode(int fd); int ei_distversion(int fd); diff --git a/lib/erl_interface/src/connect/ei_resolve.c b/lib/erl_interface/src/connect/ei_resolve.c index ba8f8fbce3..2545c5f3de 100644 --- a/lib/erl_interface/src/connect/ei_resolve.c +++ b/lib/erl_interface/src/connect/ei_resolve.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1997-2011. All Rights Reserved. + * Copyright Ericsson AB 1997-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -186,11 +186,11 @@ static int verify_dns_configuration(void) * advance: increment buf by n bytes, reduce len by same amount . */ #if defined SIZEOF_VOID_P -#define ALIGNBYTES (SIZEOF_VOID_P - 1) +#define EI_ALIGNBYTES (SIZEOF_VOID_P - 1) #else -#define ALIGNBYTES (sizeof(void*) - 1) +#define EI_ALIGNBYTES (sizeof(void*) - 1) #endif -#define align_buf(buf,len) for (;(((unsigned)buf) & ALIGNBYTES); (buf)++, len--) +#define align_buf(buf,len) for (;(((unsigned)buf) & EI_ALIGNBYTES); (buf)++, len--) #define advance_buf(buf,len,n) ((buf)+=(n),(len)-=(n)) /* "and now the tricky part..." */ @@ -282,6 +282,8 @@ static int copy_hostent(struct hostent *dest, const struct hostent *src, char *b return 0; } +#undef EI_ALIGNBYTES + /* This function is a pseudo-reentrant version of gethostbyname(). It * uses locks to serialize the call to the regular (non-reentrant) * gethostbyname() and then copies the data into the user-provided diff --git a/lib/erl_interface/src/connect/eirecv.c b/lib/erl_interface/src/connect/eirecv.c index 86852f947d..1821de3844 100644 --- a/lib/erl_interface/src/connect/eirecv.c +++ b/lib/erl_interface/src/connect/eirecv.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2010. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -108,7 +108,7 @@ ei_recv_internal (int fd, switch (msg->msgtype) { case ERL_SEND: /* { SEND, Cookie, ToPid } */ if (ei_tracelevel >= 4) show_this_msg = 1; - if (ei_decode_atom(header,&index,msg->cookie) + if (ei_decode_atom_as(header,&index,msg->cookie,sizeof(msg->cookie),ERLANG_UTF8,NULL,NULL) || ei_decode_pid(header,&index,&msg->to)) { erl_errno = EIO; @@ -120,8 +120,8 @@ ei_recv_internal (int fd, case ERL_REG_SEND: /* { REG_SEND, From, Cookie, ToName } */ if (ei_tracelevel >= 4) show_this_msg = 1; if (ei_decode_pid(header,&index,&msg->from) - || ei_decode_atom(header,&index,msg->cookie) - || ei_decode_atom(header,&index,msg->toname)) + || ei_decode_atom_as(header,&index,msg->cookie,sizeof(msg->cookie),ERLANG_UTF8,NULL,NULL) + || ei_decode_atom_as(header,&index,msg->toname,sizeof(msg->toname),ERLANG_UTF8,NULL,NULL)) { erl_errno = EIO; return -1; @@ -157,7 +157,7 @@ ei_recv_internal (int fd, case ERL_SEND_TT: /* { SEND_TT, Cookie, ToPid, TraceToken } */ if (ei_tracelevel >= 4) show_this_msg = 1; - if (ei_decode_atom(header,&index,msg->cookie) + if (ei_decode_atom_as(header,&index,msg->cookie,sizeof(msg->cookie),ERLANG_UTF8,NULL,NULL) || ei_decode_pid(header,&index,&msg->to) || ei_decode_trace(header,&index,&msg->token)) { @@ -171,8 +171,8 @@ ei_recv_internal (int fd, case ERL_REG_SEND_TT: /* { REG_SEND_TT, From, Cookie, ToName, TraceToken } */ if (ei_tracelevel >= 4) show_this_msg = 1; if (ei_decode_pid(header,&index,&msg->from) - || ei_decode_atom(header,&index,msg->cookie) - || ei_decode_atom(header,&index,msg->toname) + || ei_decode_atom_as(header,&index,msg->cookie,sizeof(msg->cookie),ERLANG_UTF8,NULL,NULL) + || ei_decode_atom_as(header,&index,msg->toname,sizeof(msg->toname),ERLANG_UTF8,NULL,NULL) || ei_decode_trace(header,&index,&msg->token)) { erl_errno = EIO; diff --git a/lib/erl_interface/src/decode/decode_atom.c b/lib/erl_interface/src/decode/decode_atom.c index c2e6a0426e..44fd4df12c 100644 --- a/lib/erl_interface/src/decode/decode_atom.c +++ b/lib/erl_interface/src/decode/decode_atom.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2011. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -21,24 +21,155 @@ #include "eiext.h" #include "putget.h" + int ei_decode_atom(const char *buf, int *index, char *p) { - const char *s = buf + *index; - const char *s0 = s; - int len; + return ei_decode_atom_as(buf, index, p, MAXATOMLEN, ERLANG_LATIN1, NULL, NULL); +} + +int ei_decode_atom_as(const char *buf, int *index, char* p, int destlen, + erlang_char_encoding want_enc, + erlang_char_encoding* was_encp, + erlang_char_encoding* res_encp) +{ + const char *s = buf + *index; + const char *s0 = s; + int len; + erlang_char_encoding got_enc; + + switch (get8(s)) { + case ERL_ATOM_EXT: + len = get16be(s); + got_enc = ERLANG_LATIN1; + break; + case ERL_SMALL_ATOM_EXT: + len = get8(s); + got_enc = ERLANG_LATIN1; + break; + case ERL_ATOM_UTF8_EXT: + len = get16be(s); + got_enc = ERLANG_UTF8; + break; + case ERL_SMALL_ATOM_UTF8_EXT: + len = get8(s); + got_enc = ERLANG_UTF8; + break; + default: + return -1; + } + + if ((want_enc & got_enc) || want_enc == ERLANG_ASCII) { + int i, found_non_ascii = 0; + if (len >= destlen) + return -1; + for (i=0; i<len; i++) { + if (s[i] & 0x80) found_non_ascii = 1; + if (p) p[i] = s[i]; + } + if (p) p[len] = 0; + if (want_enc == ERLANG_ASCII && found_non_ascii) { + return -1; + } + if (res_encp) { + *res_encp = found_non_ascii ? got_enc : ERLANG_ASCII; + } + } + else { + int plen = (got_enc == ERLANG_LATIN1) ? + latin1_to_utf8(p, s, len, destlen-1, res_encp) : + utf8_to_latin1(p, s, len, destlen-1, res_encp); + if (plen < 0) return -1; + if (p) p[plen] = 0; + } + if (was_encp) { + *was_encp = got_enc; + } + + s += len; + *index += s-s0; + return 0; +} - if (get8(s) != ERL_ATOM_EXT) return -1; - len = get16be(s); +int utf8_to_latin1(char* dst, const char* src, int slen, int destlen, + erlang_char_encoding* res_encp) +{ + const char* const dst_start = dst; + const char* const dst_end = dst + destlen; + int found_non_ascii = 0; + + while (slen > 0) { + if (dst >= dst_end) return -1; + if ((src[0] & 0x80) == 0) { + if (dst_start) { + *dst = *src; + } + ++dst; + ++src; + --slen; + } + else if (slen > 1 && + (src[0] & 0xFE) == 0xC2 && + (src[1] & 0xC0) == 0x80) { + if (dst_start) { + *dst = (char) ((src[0] << 6) | (src[1] & 0x3F)); + } + ++dst; + src += 2; + slen -= 2; + found_non_ascii = 1; + } + else return -1; + } + if (res_encp) { + *res_encp = found_non_ascii ? ERLANG_LATIN1 : ERLANG_ASCII; + } + return dst - dst_start; +} - if (len > MAXATOMLEN) return -1; +int latin1_to_utf8(char* dst, const char* src, int slen, int destlen, + erlang_char_encoding* res_encp) +{ + const char* const src_end = src + slen; + const char* const dst_start = dst; + const char* const dst_end = dst + destlen; + int found_non_ascii = 0; - if (p) { - memmove(p,s,len); - p[len] = (char)0; - } - s += len; - *index += s-s0; - - return 0; + while (src < src_end) { + if (dst >= dst_end) return -1; + if ((src[0] & 0x80) == 0) { + if (dst_start) { + *dst = *src; + } + ++dst; + } + else { + if (dst_start) { + unsigned char ch = *src; + dst[0] = 0xC0 | (ch >> 6); + dst[1] = 0x80 | (ch & 0x3F); + } + dst += 2; + found_non_ascii = 1; + } + ++src; + } + if (res_encp) { + *res_encp = found_non_ascii ? ERLANG_UTF8 : ERLANG_ASCII; + } + return dst - dst_start; } + + + +int ei_internal_get_atom(const char** bufp, char* p, + erlang_char_encoding* was_encp) +{ + int ix = 0; + if (ei_decode_atom_as(*bufp, &ix, p, MAXATOMLEN_UTF8, ERLANG_UTF8, was_encp, NULL) < 0) + return -1; + *bufp += ix; + return 0; +} + + diff --git a/lib/erl_interface/src/decode/decode_boolean.c b/lib/erl_interface/src/decode/decode_boolean.c index 9fd09c63f1..8c44a1ccfc 100644 --- a/lib/erl_interface/src/decode/decode_boolean.c +++ b/lib/erl_interface/src/decode/decode_boolean.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2009. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -24,34 +24,20 @@ /* c non-zero -> erlang "true" atom, otherwise "false" */ int ei_decode_boolean(const char *buf, int *index, int *p) { - const char *s = buf + *index; - const char *s0 = s; - int len; + char tbuf[6]; int t; - if (get8(s) != ERL_ATOM_EXT) return -1; + if (ei_decode_atom_as(buf, index, tbuf, sizeof(tbuf), ERLANG_ASCII, NULL, NULL) < 0) + return -1; - len = get16be(s); - - switch (len) { - case 4: - /* typecast makes ansi happy */ - if (strncmp((char*)s,"true",4)) return -1; - t = 1; - break; - - case 5: - if (strncmp((char*)s,"false",5)) return -1; - t = 0; - break; - - default: - return -1; - } - - s += len; + if (memcmp(tbuf, "true", 5) == 0) + t = 1; + else if (memcmp(tbuf, "false", 6) == 0) + t = 0; + else + return -1; + if (p) *p = t; - *index += s-s0; - return 0; } + diff --git a/lib/erl_interface/src/decode/decode_fun.c b/lib/erl_interface/src/decode/decode_fun.c index 64fb9e86d8..0857874c89 100644 --- a/lib/erl_interface/src/decode/decode_fun.c +++ b/lib/erl_interface/src/decode/decode_fun.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2009. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -30,6 +30,25 @@ int ei_decode_fun(const char *buf, int *index, erlang_fun *p) const char *s = buf + *index; const char *s0 = s; int i, ix, ix0, n; + erlang_pid* p_pid; + char* p_module; + erlang_char_encoding* p_module_org_enc; + long* p_index; + long* p_uniq; + long* p_old_index; + + if (p != NULL) { + p_pid = &p->pid; + p_module = &p->module[0]; + p_module_org_enc = &p->module_org_enc; + p_index = &p->index; + p_uniq = &p->uniq; + p_old_index = &p->old_index; + } + else { + p_pid = NULL; p_module = NULL; p_module_org_enc = NULL; + p_index = NULL; p_uniq = NULL; p_old_index = NULL; + } switch (get8(s)) { case ERL_FUN_EXT: @@ -39,16 +58,17 @@ int ei_decode_fun(const char *buf, int *index, erlang_fun *p) n = get32be(s); /* then the pid */ ix = 0; - if (ei_decode_pid(s, &ix, (p == NULL ? (erlang_pid*)NULL : &p->pid)) < 0) + if (ei_decode_pid(s, &ix, p_pid) < 0) return -1; /* then the module (atom) */ - if (ei_decode_atom(s, &ix, (p == NULL ? (char*)NULL : p->module)) < 0) + if (ei_decode_atom_as(s, &ix, p_module, MAXATOMLEN_UTF8, ERLANG_UTF8, + p_module_org_enc, NULL) < 0) return -1; /* then the index */ - if (ei_decode_long(s, &ix, (p == NULL ? (long*)NULL : &p->index)) < 0) + if (ei_decode_long(s, &ix, p_index) < 0) return -1; /* then the uniq */ - if (ei_decode_long(s, &ix, (p == NULL ? (long*)NULL : &p->uniq)) < 0) + if (ei_decode_long(s, &ix, p_uniq) < 0) return -1; /* finally the free vars */ ix0 = ix; @@ -84,16 +104,17 @@ int ei_decode_fun(const char *buf, int *index, erlang_fun *p) if (p != NULL) p->n_free_vars = i; /* then the module (atom) */ ix = 0; - if (ei_decode_atom(s, &ix, (p == NULL ? (char*)NULL : p->module)) < 0) + if (ei_decode_atom_as(s, &ix, p_module, MAXATOMLEN_UTF8, ERLANG_UTF8, + p_module_org_enc, NULL) < 0) return -1; /* then the old_index */ - if (ei_decode_long(s, &ix, (p == NULL ? (long*)NULL : &p->old_index)) < 0) + if (ei_decode_long(s, &ix, p_old_index) < 0) return -1; /* then the old_uniq */ - if (ei_decode_long(s, &ix, (p == NULL ? (long*)NULL : &p->uniq)) < 0) + if (ei_decode_long(s, &ix, p_uniq) < 0) return -1; /* the the pid */ - if (ei_decode_pid(s, &ix, (p == NULL ? (erlang_pid*)NULL : &p->pid)) < 0) + if (ei_decode_pid(s, &ix, p_pid) < 0) return -1; /* finally the free vars */ s += ix; diff --git a/lib/erl_interface/src/decode/decode_pid.c b/lib/erl_interface/src/decode/decode_pid.c index 9ed1c36db6..d429fb2fd8 100644 --- a/lib/erl_interface/src/decode/decode_pid.c +++ b/lib/erl_interface/src/decode/decode_pid.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2011. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -21,34 +21,24 @@ #include "eiext.h" #include "putget.h" + int ei_decode_pid(const char *buf, int *index, erlang_pid *p) { const char *s = buf + *index; const char *s0 = s; - int len; if (get8(s) != ERL_PID_EXT) return -1; - /* first the nodename */ - if (get8(s) != ERL_ATOM_EXT) return -1; - - len = get16be(s); - - if (len > MAXATOMLEN) return -1; - - if (p) { - memmove(p->node, s, len); - p->node[len] = (char)0; - } - s += len; - - /* now the numbers: num (4), serial (4), creation (1) */ if (p) { + if (get_atom(&s, p->node, &p->node_org_enc) < 0) return -1; p->num = get32be(s) & 0x7fff; /* 15 bits */ p->serial = get32be(s) & 0x1fff; /* 13 bits */ p->creation = get8(s) & 0x03; /* 2 bits */ } - else s+= 9; + else { + if (get_atom(&s, NULL, NULL) < 0) return -1; + s+= 9; + } *index += s-s0; diff --git a/lib/erl_interface/src/decode/decode_port.c b/lib/erl_interface/src/decode/decode_port.c index 28abed801a..7a691f0be6 100644 --- a/lib/erl_interface/src/decode/decode_port.c +++ b/lib/erl_interface/src/decode/decode_port.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2011. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -25,29 +25,18 @@ int ei_decode_port(const char *buf, int *index, erlang_port *p) { const char *s = buf + *index; const char *s0 = s; - int len; if (get8(s) != ERL_PORT_EXT) return -1; - /* first the nodename */ - if (get8(s) != ERL_ATOM_EXT) return -1; - - len = get16be(s); - - if (len > MAXATOMLEN) return -1; - - if (p) { - memmove(p->node, s, len); - p->node[len] = (char)0; - } - s += len; - - /* now the numbers: num (4), creation (1) */ if (p) { + if (get_atom(&s, p->node, &p->node_org_enc) < 0) return -1; p->id = get32be(s) & 0x0fffffff /* 28 bits */; p->creation = get8(s) & 0x03; } - else s += 5; + else { + if (get_atom(&s, NULL, NULL) < 0) return -1; + s += 5; + } *index += s-s0; diff --git a/lib/erl_interface/src/decode/decode_ref.c b/lib/erl_interface/src/decode/decode_ref.c index 7b15808bc5..01e3061cb4 100644 --- a/lib/erl_interface/src/decode/decode_ref.c +++ b/lib/erl_interface/src/decode/decode_ref.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2011. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -21,35 +21,25 @@ #include "eiext.h" #include "putget.h" + int ei_decode_ref(const char *buf, int *index, erlang_ref *p) { const char *s = buf + *index; const char *s0 = s; - int count, len, i; + int count, i; switch (get8(s)) { case ERL_REFERENCE_EXT: - - /* first the nodename */ - if (get8(s) != ERL_ATOM_EXT) return -1; - - len = get16be(s); - - if (len > MAXATOMLEN) return -1; - - if (p) { - memmove(p->node, s, len); - p->node[len] = (char)0; - } - s += len; - - /* now the numbers: num (4), creation (1) */ if (p) { + if (get_atom(&s, p->node, &p->node_org_enc) < 0) return -1; p->n[0] = get32be(s); p->len = 1; p->creation = get8(s) & 0x03; } - else s += 5; + else { + if (get_atom(&s, NULL, NULL) < 0) return -1; + s += 5; + } *index += s-s0; @@ -59,24 +49,16 @@ int ei_decode_ref(const char *buf, int *index, erlang_ref *p) case ERL_NEW_REFERENCE_EXT: /* first the integer count */ count = get16be(s); - if (p) p->len = count; - - /* then the nodename */ - if (get8(s) != ERL_ATOM_EXT) return -1; - len = get16be(s); - if (len > MAXATOMLEN) return -1; if (p) { - memmove(p->node, s, len); - p->node[len] = (char)0; - } - s += len; - - /* creation */ - if (p) { + p->len = count; + if (get_atom(&s, p->node, &p->node_org_enc) < 0) return -1; p->creation = get8(s) & 0x03; } - else s += 1; + else { + if (get_atom(&s, NULL, NULL) < 0) return -1; + s += 1; + } /* finally the id integers */ if (p) { @@ -95,3 +77,4 @@ int ei_decode_ref(const char *buf, int *index, erlang_ref *p) return -1; } } + diff --git a/lib/erl_interface/src/decode/decode_skip.c b/lib/erl_interface/src/decode/decode_skip.c index f6c5d861ab..e2bfe1f802 100644 --- a/lib/erl_interface/src/decode/decode_skip.c +++ b/lib/erl_interface/src/decode/decode_skip.c @@ -30,7 +30,8 @@ int ei_skip_term(const char* buf, int* index) switch (ty) { case ERL_ATOM_EXT: /* FIXME: what if some weird locale is in use? */ - if (ei_decode_atom(buf, index, NULL) < 0) return -1; + if (ei_decode_atom_as(buf, index, NULL, MAXATOMLEN_UTF8, (ERLANG_LATIN1|ERLANG_UTF8), + NULL, NULL) < 0) return -1; break; case ERL_PID_EXT: if (ei_decode_pid(buf, index, NULL) < 0) return -1; diff --git a/lib/erl_interface/src/decode/decode_trace.c b/lib/erl_interface/src/decode/decode_trace.c index ebaa78e29e..88fb3451ec 100644 --- a/lib/erl_interface/src/decode/decode_trace.c +++ b/lib/erl_interface/src/decode/decode_trace.c @@ -22,18 +22,30 @@ int ei_decode_trace(const char *buf, int *index, erlang_trace *p) { int arity = 0; - int tindex = *index; - - /* use a temporary index if any function should fail */ + int tindex = *index; /* use a temporary index if any function should fail */ + long *p_flags, *p_label, *p_serial, *p_prev; + erlang_pid *p_from; + + if (p != NULL) { + p_flags = &p->flags; + p_label = &p->label; + p_serial = &p->serial; + p_prev = &p->prev; + p_from = &p->from; + } + else { + p_flags = p_label = p_serial = p_prev = NULL; + p_from = NULL; + } /* { Flags, Label, Serial, FromPid, Prev } */ if (ei_decode_tuple_header(buf, &tindex, &arity) || (arity != 5) - || ei_decode_long(buf, &tindex, &p->flags) - || ei_decode_long(buf, &tindex, &p->label) - || ei_decode_long(buf, &tindex, &p->serial) - || ei_decode_pid( buf, &tindex, &p->from) - || ei_decode_long(buf, &tindex, &p->prev)) return -1; + || ei_decode_long(buf, &tindex, p_flags) + || ei_decode_long(buf, &tindex, p_label) + || ei_decode_long(buf, &tindex, p_serial) + || ei_decode_pid( buf, &tindex, p_from) + || ei_decode_long(buf, &tindex, p_prev)) return -1; /* index is updated by the functions we called */ diff --git a/lib/erl_interface/src/encode/encode_atom.c b/lib/erl_interface/src/encode/encode_atom.c index 6f41f045e0..df4b0af5db 100644 --- a/lib/erl_interface/src/encode/encode_atom.c +++ b/lib/erl_interface/src/encode/encode_atom.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2011. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -22,29 +22,112 @@ #include "eiext.h" #include "putget.h" + +static int verify_ascii_atom(const char* src, int slen); +static int verify_utf8_atom(const char* src, int slen); + + int ei_encode_atom(char *buf, int *index, const char *p) { size_t len = strlen(p); - if (len >= INT_MAX) return -1; - return ei_encode_atom_len(buf, index, p, len); + if (len >= MAXATOMLEN) + len = MAXATOMLEN - 1; + return ei_encode_atom_len_as(buf, index, p, len, ERLANG_LATIN1, ERLANG_LATIN1); } int ei_encode_atom_len(char *buf, int *index, const char *p, int len) { + /* This function is documented to truncate at MAXATOMLEN (256) */ + if (len >= MAXATOMLEN) + len = MAXATOMLEN - 1; + return ei_encode_atom_len_as(buf, index, p, len, ERLANG_LATIN1, ERLANG_LATIN1); +} + +int ei_encode_atom_as(char *buf, int *index, const char *p, + erlang_char_encoding from_enc, + erlang_char_encoding to_enc) +{ + return ei_encode_atom_len_as(buf, index, p, strlen(p), from_enc, to_enc); +} + +int ei_encode_atom_len_as(char *buf, int *index, const char *p, int len, + erlang_char_encoding from_enc, + erlang_char_encoding to_enc) +{ char *s = buf + *index; char *s0 = s; + int offs; - /* This function is documented to truncate at MAXATOMLEN (256) */ - if (len > MAXATOMLEN) - len = MAXATOMLEN; + if (len >= MAXATOMLEN && (from_enc & (ERLANG_LATIN1|ERLANG_ASCII))) { + return -1; + } - if (!buf) s += 3; - else { - put8(s,ERL_ATOM_EXT); - put16be(s,len); + switch(to_enc) { + case ERLANG_LATIN1: + if (buf) { + put8(s,ERL_ATOM_EXT); + switch (from_enc) { + case ERLANG_UTF8: + len = utf8_to_latin1(s+2, p, len, MAXATOMLEN-1, NULL); + if (len < 0) return -1; + break; + case ERLANG_ASCII: + if (verify_ascii_atom(p, len) < 0) return -1; + memcpy(s+2, p, len); + break; + case ERLANG_LATIN1: + memcpy(s+2, p, len); + break; + default: + return -1; + } + put16be(s,len); + } + else { + s += 3; + if (from_enc == ERLANG_UTF8) { + len = utf8_to_latin1(NULL, p, len, MAXATOMLEN-1, NULL); + if (len < 0) return -1; + } else if (from_enc == ERLANG_ASCII) + if (verify_ascii_atom(p, len) < 0) return -1; + } + break; + + case ERLANG_UTF8: + offs = 1 + 1; + switch (from_enc) { + case ERLANG_LATIN1: + if (len >= 256/2) offs++; + len = latin1_to_utf8((buf ? s+offs : NULL), p, len, MAXATOMLEN_UTF8-1, NULL); + break; + case ERLANG_ASCII: + if (verify_ascii_atom(p, len) < 0) return -1; + if (buf) memcpy(s+offs,p,len); + break; + case ERLANG_UTF8: + if (len >= 256) offs++; + if (verify_utf8_atom(p, len) < 0) return -1; + if (buf) memcpy(s+offs,p,len); + break; + default: + return -1; + } + if (buf) { + if (offs == 2) { + put8(s, ERL_SMALL_ATOM_UTF8_EXT); + put8(s, len); + } + else { + put8(s, ERL_ATOM_UTF8_EXT); + put16be(s, len); + } + } + else s+= offs; + break; - memmove(s,p,len); /* unterminated string */ + default: + return -1; } s += len; @@ -53,3 +136,55 @@ int ei_encode_atom_len(char *buf, int *index, const char *p, int len) return 0; } +int +ei_internal_put_atom(char** bufp, const char* p, int slen, + erlang_char_encoding to_enc) +{ + int ix = 0; + if (ei_encode_atom_len_as(*bufp, &ix, p, slen, ERLANG_UTF8, to_enc) < 0) + return -1; + *bufp += ix; + return 0; +} + + +int verify_ascii_atom(const char* src, int slen) +{ + while (slen > 0) { + if ((src[0] & 0x80) != 0) return -1; + src++; + slen--; + } + return 0; +} + +int verify_utf8_atom(const char* src, int slen) +{ + int num_chars = 0; + + while (slen > 0) { + if (++num_chars >= MAXATOMLEN) return -1; + if ((src[0] & 0x80) != 0) { + if ((src[0] & 0xE0) == 0xC0) { + if (slen < 2 || (src[1] & 0xC0) != 0x80) return -1; + src++; + slen--; + } + else if ((src[0] & 0xF0) == 0xE0) { + if (slen < 3 || (src[1] & 0xC0) != 0x80 || (src[2] & 0xC0) != 0x80) return -1; + src += 2; + slen -= 2; + } + else if ((src[0] & 0xF8) == 0xF0) { + if (slen < 4 || (src[1] & 0xC0) != 0x80 || (src[2] & 0xC0) != 0x80 || (src[3] & 0xC0) != 0x80) return -1; + src += 3; + slen -= 3; + } + else return -1; + } + src++; + slen--; + } + return 0; +} + diff --git a/lib/erl_interface/src/encode/encode_fun.c b/lib/erl_interface/src/encode/encode_fun.c index 54ee2083d6..455d8a5a94 100644 --- a/lib/erl_interface/src/encode/encode_fun.c +++ b/lib/erl_interface/src/encode/encode_fun.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2009. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -35,7 +35,7 @@ int ei_encode_fun(char *buf, int *index, const erlang_fun *p) ix += sizeof(char) + 4; if (ei_encode_pid(buf, &ix, &p->pid) < 0) return -1; - if (ei_encode_atom(buf, &ix, p->module) < 0) + if (ei_encode_atom_as(buf, &ix, p->module, ERLANG_UTF8, p->module_org_enc) < 0) return -1; if (ei_encode_long(buf, &ix, p->index) < 0) return -1; @@ -60,7 +60,7 @@ int ei_encode_fun(char *buf, int *index, const erlang_fun *p) } else size_p = NULL; ix += 1 + 4 + 1 + sizeof(p->md5) + 4 + 4; - if (ei_encode_atom(buf, &ix, p->module) < 0) + if (ei_encode_atom_as(buf, &ix, p->module, ERLANG_UTF8, p->module_org_enc) < 0) return -1; if (ei_encode_long(buf, &ix, p->old_index) < 0) return -1; diff --git a/lib/erl_interface/src/encode/encode_pid.c b/lib/erl_interface/src/encode/encode_pid.c index ee7f235c17..903c9cce00 100644 --- a/lib/erl_interface/src/encode/encode_pid.c +++ b/lib/erl_interface/src/encode/encode_pid.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2009. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -24,29 +24,23 @@ int ei_encode_pid(char *buf, int *index, const erlang_pid *p) { char *s = buf + *index; - char *s0 = s; - int len = strlen(p->node); - - if (!buf) s += 13 + len; - else { - put8(s,ERL_PID_EXT); - /* first the nodename */ - put8(s,ERL_ATOM_EXT); + ++(*index); /* skip ERL_PID_EXT */ + if (ei_encode_atom_len_as(buf, index, p->node, strlen(p->node), ERLANG_UTF8, p->node_org_enc) < 0) + return -1; + + if (buf) { + put8(s,ERL_PID_EXT); - put16be(s,len); - - memmove(s, p->node, len); - s += len; + s = buf + *index; /* now the integers */ put32be(s,p->num & 0x7fff); /* 15 bits */ put32be(s,p->serial & 0x1fff); /* 13 bits */ put8(s,(p->creation & 0x03)); /* 2 bits */ } - - *index += s-s0; - + + *index += 4 + 4 + 1; return 0; } diff --git a/lib/erl_interface/src/encode/encode_port.c b/lib/erl_interface/src/encode/encode_port.c index fbbb33182e..c729aeb4eb 100644 --- a/lib/erl_interface/src/encode/encode_port.c +++ b/lib/erl_interface/src/encode/encode_port.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2009. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -24,28 +24,23 @@ int ei_encode_port(char *buf, int *index, const erlang_port *p) { char *s = buf + *index; - char *s0 = s; - int len = strlen(p->node); - - if (!buf) s += 9 + len; - else { - put8(s,ERL_PORT_EXT); - /* first the nodename */ - put8(s,ERL_ATOM_EXT); + ++(*index); /* skip ERL_PORT_EXT */ + if (ei_encode_atom_len_as(buf, index, p->node, strlen(p->node), ERLANG_UTF8, + p->node_org_enc) < 0) { + return -1; + } + if (buf) { + put8(s,ERL_PORT_EXT); - put16be(s,len); - - memmove(s, p->node, len); - s += len; + s = buf + *index; /* now the integers */ put32be(s,p->id & 0x0fffffff /* 28 bits */); put8(s,(p->creation & 0x03)); } - *index += s-s0; - + *index += 4 + 1; return 0; } diff --git a/lib/erl_interface/src/encode/encode_ref.c b/lib/erl_interface/src/encode/encode_ref.c index 292b452864..3511366bef 100644 --- a/lib/erl_interface/src/encode/encode_ref.c +++ b/lib/erl_interface/src/encode/encode_ref.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2009. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -24,36 +24,32 @@ int ei_encode_ref(char *buf, int *index, const erlang_ref *p) { char *s = buf + *index; - char *s0 = s; - int len = strlen(p->node); int i; + (*index) += 1 + 2; /* skip to node atom */ + if (ei_encode_atom_len_as(buf, index, p->node, strlen(p->node), ERLANG_UTF8, + p->node_org_enc) < 0) { + return -1; + } + /* Always encode as an extended reference; all participating parties are now expected to be able to decode extended references. */ - if (!buf) s += 1 + 2 + (3+len) + p->len*4 + 1; - else { + if (buf) { put8(s,ERL_NEW_REFERENCE_EXT); /* first, number of integers */ put16be(s, p->len); /* then the nodename */ - put8(s,ERL_ATOM_EXT); - - put16be(s,len); - - memmove(s, p->node, len); - s += len; + s = buf + *index; /* now the integers */ put8(s,(p->creation & 0x03)); for (i = 0; i < p->len; i++) put32be(s,p->n[i]); - - } - - *index += s-s0; + } + *index += p->len*4 + 1; return 0; } diff --git a/lib/erl_interface/src/legacy/erl_connect.c b/lib/erl_interface/src/legacy/erl_connect.c index 41d4fa3138..eca16497dc 100644 --- a/lib/erl_interface/src/legacy/erl_connect.c +++ b/lib/erl_interface/src/legacy/erl_connect.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2011. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -125,7 +125,7 @@ static ei_cnode erl_if_ec; int erl_connect_init(int this_node_number, char *cookie, short creation) { - char nn[MAXATOMLEN+1]; + char nn[MAXATOMLEN]; sprintf(nn, "c%d", this_node_number); @@ -247,9 +247,15 @@ int erl_send(int fd, ETERM *to ,ETERM *msg) erl_errno = EINVAL; return -1; } - - strncpy(topid.node, (char *)ERL_PID_NODE(to), sizeof(topid.node)); - topid.node[sizeof(topid.node)-1] = '\0'; + + if (to->uval.pidval.node.latin1) { + strcpy(topid.node, to->uval.pidval.node.latin1); + topid.node_org_enc = ERLANG_LATIN1; + } + else { + strcpy(topid.node, to->uval.pidval.node.utf8); + topid.node_org_enc = ERLANG_UTF8; + } topid.num = ERL_PID_NUMBER(to); topid.serial = ERL_PID_SERIAL(to); topid.creation = ERL_PID_CREATION(to); @@ -263,7 +269,7 @@ static int erl_do_receive_msg(int fd, ei_x_buff* x, ErlMessage* emsg) erlang_msg msg; int r; - msg.from.node[0] = msg.to.node[0] = '\0'; + msg.from.node[0] = msg.to.node[0] = msg.toname[0] = '\0'; r = ei_do_receive_msg(fd, 0, &msg, x, 0); if (r == ERL_MSG) { @@ -299,7 +305,7 @@ static int erl_do_receive_msg(int fd, ei_x_buff* x, ErlMessage* emsg) emsg->to = erl_mk_pid(msg.to.node, msg.to.num, msg.to.serial, msg.to.creation); else emsg->to = NULL; - memcpy(emsg->to_name, msg.toname, MAXATOMLEN+1); + strcpy(emsg->to_name, msg.toname); return r; } diff --git a/lib/erl_interface/src/legacy/erl_eterm.c b/lib/erl_interface/src/legacy/erl_eterm.c index 8d559f0f55..7ca4f430de 100644 --- a/lib/erl_interface/src/legacy/erl_eterm.c +++ b/lib/erl_interface/src/legacy/erl_eterm.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2010. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -36,6 +36,7 @@ #include "erl_error.h" #include "erl_internal.h" #include "ei_internal.h" +#include "putget.h" #define ERL_IS_BYTE(x) (ERL_IS_INTEGER(x) && (ERL_INT_VALUE(x) & ~0xFF) == 0) @@ -142,9 +143,7 @@ ETERM *erl_mk_atom (const char *s) ep = erl_alloc_eterm(ERL_ATOM); ERL_COUNT(ep) = 1; - ERL_ATOM_SIZE(ep) = strlen(s); - if ((ERL_ATOM_PTR(ep) = strsave(s)) == NULL) - { + if (erl_atom_init_latin1(&ep->uval.aval.d, s) == NULL) { erl_free_term(ep); erl_errno = ENOMEM; return NULL; @@ -152,6 +151,65 @@ ETERM *erl_mk_atom (const char *s) return ep; } +char* erl_atom_ptr_latin1(Erl_Atom_data* a) +{ + if (a->latin1 == NULL) { + erlang_char_encoding enc; + a->lenL = utf8_to_latin1(NULL, a->utf8, a->lenU, a->lenU, &enc); + if (a->lenL < 0) { + a->lenL = 0; + return NULL; + } + if (enc == ERLANG_ASCII) { + a->latin1 = a->utf8; + } + else { + a->latin1 = malloc(a->lenL+1); + utf8_to_latin1(a->latin1, a->utf8, a->lenU, a->lenL, NULL); + a->latin1[a->lenL] = '\0'; + } + } + return a->latin1; +} + +char* erl_atom_ptr_utf8(Erl_Atom_data* a) +{ + if (a->utf8 == NULL) { + int dlen = a->lenL * 2; /* over estimation */ + a->utf8 = malloc(dlen + 1); + a->lenU = latin1_to_utf8(a->utf8, a->latin1, a->lenL, dlen, NULL); + a->utf8[a->lenU] = '\0'; + } + return a->utf8; + +} +int erl_atom_size_latin1(Erl_Atom_data* a) +{ + if (a->latin1 == NULL) { + erl_atom_ptr_latin1(a); + } + return a->lenL; +} +int erl_atom_size_utf8(Erl_Atom_data* a) +{ + if (a->utf8 == NULL) { + erl_atom_ptr_utf8(a); + } + return a->lenU; +} +char* erl_atom_init_latin1(Erl_Atom_data* a, const char* s) +{ + a->lenL = strlen(s); + if ((a->latin1 = strsave(s)) == NULL) + { + return NULL; + } + a->utf8 = NULL; + a->lenU = 0; + return a->latin1; +} + + /* * Given a string as input, creates a list. */ @@ -208,12 +266,19 @@ ETERM *erl_mk_pid(const char *node, ep = erl_alloc_eterm(ERL_PID); ERL_COUNT(ep) = 1; - if ((ERL_PID_NODE(ep) = strsave(node)) == NULL) + if (erl_atom_init_latin1(&ep->uval.pidval.node, node) == NULL) { erl_free_term(ep); erl_errno = ENOMEM; return NULL; } + erl_mk_pid_helper(ep, number, serial, creation); + return ep; +} + +void erl_mk_pid_helper(ETERM *ep, unsigned int number, + unsigned int serial, unsigned char creation) +{ ERL_PID_NUMBER(ep) = number & 0x7fff; /* 15 bits */ if (ei_internal_use_r9_pids_ports()) { ERL_PID_SERIAL(ep) = serial & 0x07; /* 3 bits */ @@ -222,7 +287,6 @@ ETERM *erl_mk_pid(const char *node, ERL_PID_SERIAL(ep) = serial & 0x1fff; /* 13 bits */ } ERL_PID_CREATION(ep) = creation & 0x03; /* 2 bits */ - return ep; } /* @@ -239,12 +303,18 @@ ETERM *erl_mk_port(const char *node, ep = erl_alloc_eterm(ERL_PORT); ERL_COUNT(ep) = 1; - if ((ERL_PORT_NODE(ep) = strsave(node)) == NULL) + if (erl_atom_init_latin1(&ep->uval.portval.node, node) == NULL) { erl_free_term(ep); erl_errno = ENOMEM; return NULL; } + erl_mk_port_helper(ep, number, creation); + return ep; +} + +void erl_mk_port_helper(ETERM* ep, unsigned number, unsigned char creation) +{ if (ei_internal_use_r9_pids_ports()) { ERL_PORT_NUMBER(ep) = number & 0x3ffff; /* 18 bits */ } @@ -252,29 +322,29 @@ ETERM *erl_mk_port(const char *node, ERL_PORT_NUMBER(ep) = number & 0x0fffffff; /* 18 bits */ } ERL_PORT_CREATION(ep) = creation & 0x03; /* 2 bits */ - return ep; } /* * Create any kind of reference. */ -ETERM *__erl_mk_reference (const char *node, +ETERM *__erl_mk_reference (ETERM* t, + const char *node, size_t len, unsigned int n[], unsigned char creation) { - ETERM * t; - - if (node == NULL) return NULL; - - t = erl_alloc_eterm(ERL_REF); - ERL_COUNT(t) = 1; - - if ((ERL_REF_NODE(t) = strsave(node)) == NULL) - { - erl_free_term(t); - erl_errno = ENOMEM; - return NULL; + if (t == NULL) { + if (node == NULL) return NULL; + + t = erl_alloc_eterm(ERL_REF); + ERL_COUNT(t) = 1; + + if (erl_atom_init_latin1(&t->uval.refval.node, node) == NULL) + { + erl_free_term(t); + erl_errno = ENOMEM; + return NULL; + } } ERL_REF_LEN(t) = len; ERL_REF_NUMBERS(t)[0] = n[0] & 0x3ffff; /* 18 bits */ @@ -294,7 +364,7 @@ ETERM *erl_mk_ref (const char *node, { unsigned int n[3] = {0, 0, 0}; n[0] = number; - return __erl_mk_reference(node, 1, n, creation); + return __erl_mk_reference(NULL, node, 1, n, creation); } /* @@ -307,7 +377,7 @@ erl_mk_long_ref (const char *node, { unsigned int n[3] = {0, 0, 0}; n[0] = n3; n[1] = n2; n[2] = n1; - return __erl_mk_reference(node, 3, n, creation); + return __erl_mk_reference(NULL, node, 3, n, creation); } /* @@ -758,6 +828,28 @@ int erl_iolist_length (const ETERM* term) return -1; } +static int erl_atom_copy(Erl_Atom_data* dst, const Erl_Atom_data* src) +{ + if (src->latin1 == src->utf8) { + dst->latin1 = dst->utf8 = strsave(src->latin1); + dst->lenL = dst->lenU = strlen(src->latin1); + } + else if (src->latin1) { + dst->latin1 = strsave(src->latin1); + dst->lenL = strlen(src->latin1); + dst->utf8 = NULL; + dst->lenU = 0; + } + else { + dst->utf8 = strsave(src->utf8); + dst->lenU = strlen(src->utf8); + dst->latin1 = NULL; + dst->lenL = 0; + } + return (dst->latin1 != NULL || dst->utf8 == NULL); +} + + /* * Return a brand NEW COPY of an ETERM. */ @@ -796,9 +888,7 @@ ETERM *erl_copy_term(const ETERM *ep) ERL_FLOAT_VALUE(cp) = ERL_FLOAT_VALUE(ep); break; case ERL_ATOM: - ERL_ATOM_SIZE(cp) = ERL_ATOM_SIZE(ep); - ERL_ATOM_PTR(cp) = strsave(ERL_ATOM_PTR(ep)); - if (ERL_ATOM_PTR(cp) == NULL) + if (!erl_atom_copy(&cp->uval.aval.d, &ep->uval.aval.d)) { erl_free_term(cp); erl_errno = ENOMEM; @@ -810,17 +900,17 @@ ETERM *erl_copy_term(const ETERM *ep) name and plug in. Somewhat ugly (also done with port and ref below). */ memcpy(&cp->uval.pidval, &ep->uval.pidval, sizeof(Erl_Pid)); - ERL_PID_NODE(cp) = strsave(ERL_PID_NODE(ep)); + erl_atom_copy(&cp->uval.pidval.node, &ep->uval.pidval.node); ERL_COUNT(cp) = 1; break; case ERL_PORT: memcpy(&cp->uval.portval, &ep->uval.portval, sizeof(Erl_Port)); - ERL_PORT_NODE(cp) = strsave(ERL_PORT_NODE(ep)); + erl_atom_copy(&cp->uval.portval.node, &ep->uval.portval.node); ERL_COUNT(cp) = 1; break; case ERL_REF: memcpy(&cp->uval.refval, &ep->uval.refval, sizeof(Erl_Ref)); - ERL_REF_NODE(cp) = strsave(ERL_REF_NODE(ep)); + erl_atom_copy(&cp->uval.refval.node, &ep->uval.refval.node); ERL_COUNT(cp) = 1; break; case ERL_LIST: @@ -883,29 +973,29 @@ int erl_print_term(FILE *fp, const ETERM *ep) j = i = doquote = 0; switch(ERL_TYPE(ep)) { - case ERL_ATOM: + case ERL_ATOM: { + char* adata = ERL_ATOM_PTR(ep); /* FIXME: what if some weird locale is in use? */ - if (!islower((int)ERL_ATOM_PTR(ep)[0])) + if (!islower(adata[0])) doquote = 1; for (i = 0; !doquote && i < ERL_ATOM_SIZE(ep); i++) { - doquote = !(isalnum((int)ERL_ATOM_PTR(ep)[i]) - || (ERL_ATOM_PTR(ep)[i] == '_')); + doquote = !(isalnum(adata[i]) || (adata[i] == '_')); } if (doquote) { putc('\'', fp); ch_written++; } - fputs(ERL_ATOM_PTR(ep), fp); + fputs(adata, fp); ch_written += ERL_ATOM_SIZE(ep); if (doquote) { putc('\'', fp); ch_written++; } break; - + } case ERL_VARIABLE: if (!isupper((int)ERL_VAR_NAME(ep)[0])) { doquote = 1; diff --git a/lib/erl_interface/src/legacy/erl_eterm.h b/lib/erl_interface/src/legacy/erl_eterm.h index 41b008f04f..ca4ec7925a 100644 --- a/lib/erl_interface/src/legacy/erl_eterm.h +++ b/lib/erl_interface/src/legacy/erl_eterm.h @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2009. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -55,7 +55,9 @@ typedef struct _heapmark { } Erl_HeapMark; -ETERM * __erl_mk_reference(const char *, size_t, unsigned int n[], unsigned char); +void erl_mk_port_helper(ETERM* ep, unsigned number, unsigned char creation); +void erl_mk_pid_helper(ETERM*, unsigned,unsigned, unsigned char); +ETERM * __erl_mk_reference(ETERM*, const char *, size_t, unsigned int n[], unsigned char); int erl_current_fix_desc(void); #endif /* _ERL_ETERM_H */ diff --git a/lib/erl_interface/src/legacy/erl_format.c b/lib/erl_interface/src/legacy/erl_format.c index dc85806c36..fda05cbdb8 100644 --- a/lib/erl_interface/src/legacy/erl_format.c +++ b/lib/erl_interface/src/legacy/erl_format.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2011. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -574,10 +574,22 @@ static int ematch(ETERM *p, ETERM *t) switch (type_p) { - case ERL_ATOM: - return p->uval.aval.len == t->uval.aval.len && - memcmp(p->uval.aval.a, t->uval.aval.a, p->uval.aval.len) == 0; - + case ERL_ATOM: { + Erl_Atom_data* pa = &p->uval.aval.d; + Erl_Atom_data* ta = &t->uval.aval.d; + if (pa->utf8 && ta->utf8) { + return pa->lenU == ta->lenU && memcmp(pa->utf8, ta->utf8, pa->lenU)==0; + } + else if (pa->latin1 && ta->latin1) { + return pa->lenL == ta->lenL && memcmp(pa->latin1, ta->latin1, pa->lenL)==0; + } + else if (pa->latin1) { + return cmp_latin1_vs_utf8(pa->latin1, pa->lenL, ta->utf8, ta->lenU)==0; + } + else { + return cmp_latin1_vs_utf8(ta->latin1, ta->lenL, pa->utf8, pa->lenU)==0; + } + } case ERL_VARIABLE: if (strcmp(p->uval.vval.name, "_") == 0) /* anon. variable */ return ERL_TRUE; diff --git a/lib/erl_interface/src/legacy/erl_malloc.c b/lib/erl_interface/src/legacy/erl_malloc.c index f51a6c69b3..c320a00bfe 100644 --- a/lib/erl_interface/src/legacy/erl_malloc.c +++ b/lib/erl_interface/src/legacy/erl_malloc.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2009. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -112,6 +112,18 @@ do { \ (ptr) = NULL; \ } while (0) +static void erl_atom_free(Erl_Atom_data* p) +{ + erl_free(p->latin1); + if (p->utf8 != p->latin1) { + erl_free(p->utf8); + } + p->latin1 = NULL; + p->utf8 = NULL; + p->lenL = 0; + p->lenU = 0; +} + static void _erl_free_term (ETERM *ep, int external, int compound) { restart: @@ -122,7 +134,7 @@ restart: switch(ERL_TYPE(ep)) { case ERL_ATOM: - FREE_AND_CLEAR(ERL_ATOM_PTR(ep)); + erl_atom_free(&ep->uval.aval.d); break; case ERL_VARIABLE: FREE_AND_CLEAR(ERL_VAR_NAME(ep)); @@ -161,13 +173,13 @@ restart: FREE_AND_CLEAR(ERL_BIN_PTR(ep)); break; case ERL_PID: - FREE_AND_CLEAR(ERL_PID_NODE(ep)); + erl_atom_free(&ep->uval.pidval.node); break; case ERL_PORT: - FREE_AND_CLEAR(ERL_PORT_NODE(ep)); + erl_atom_free(&ep->uval.portval.node); break; case ERL_REF: - FREE_AND_CLEAR(ERL_REF_NODE(ep)); + erl_atom_free(&ep->uval.refval.node); break; case ERL_EMPTY_LIST: case ERL_INTEGER: diff --git a/lib/erl_interface/src/legacy/erl_marshal.c b/lib/erl_interface/src/legacy/erl_marshal.c index dad715c762..e207b5f0f1 100644 --- a/lib/erl_interface/src/legacy/erl_marshal.c +++ b/lib/erl_interface/src/legacy/erl_marshal.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1996-2011. All Rights Reserved. + * Copyright Ericsson AB 1996-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -44,6 +44,9 @@ int erl_fp_compare(unsigned *a, unsigned *b); static void erl_long_to_fp(long l, unsigned *d); #endif +static int cmpbytes(unsigned char* s1,int l1,unsigned char* s2,int l2); +static int cmpatoms(unsigned char* s1, int l1, unsigned char tag1, unsigned char* s2, int l2, unsigned char tag2); + /* Used when comparing two encoded byte arrays */ /* this global data is ok (from threading point of view) since it is * initialized once and never changed @@ -51,7 +54,13 @@ static void erl_long_to_fp(long l, unsigned *d); #define CMP_ARRAY_SIZE 256 /* FIXME problem for threaded ? */ -static char cmp_array[CMP_ARRAY_SIZE]; + +static enum +{ + ERL_NUM_CMP=1, ERL_ATOM_CMP, ERL_REF_CMP, ERL_FUN_CMP, ERL_PORT_CMP, + ERL_PID_CMP, ERL_TUPLE_CMP, ERL_NIL_CMP, ERL_LIST_CMP, ERL_BIN_CMP +}cmp_array[CMP_ARRAY_SIZE]; + static int init_cmp_array_p=1; /* initialize array, the first time */ #if defined(VXWORKS) && CPU == PPC860 @@ -69,10 +78,8 @@ static int init_cmp_array_p=1; /* initialize array, the first time */ static int cmp_floats(double f1, double f2); static INLINE double to_float(long l); -#define ERL_NUM_CMP 1 -#define ERL_REF_CMP 3 - #define IS_ERL_NUM(t) (cmp_array[t]==ERL_NUM_CMP) +#define IS_ERL_ATOM(t) (cmp_array[t]==ERL_ATOM_CMP) #define CMP_NUM_CLASS_SIZE 256 static unsigned char cmp_num_class[CMP_NUM_CLASS_SIZE]; @@ -100,25 +107,28 @@ void erl_init_marshal(void) { if (init_cmp_array_p) { memset(cmp_array, 0, CMP_ARRAY_SIZE); - cmp_array[ERL_SMALL_INTEGER_EXT] = 1; - cmp_array[ERL_INTEGER_EXT] = 1; - cmp_array[ERL_FLOAT_EXT] = 1; - cmp_array[NEW_FLOAT_EXT] = 1; - cmp_array[ERL_SMALL_BIG_EXT] = 1; - cmp_array[ERL_LARGE_BIG_EXT] = 1; - cmp_array[ERL_ATOM_EXT] = 2; - cmp_array[ERL_REFERENCE_EXT] = 3; - cmp_array[ERL_NEW_REFERENCE_EXT] = 3; - cmp_array[ERL_FUN_EXT] = 4; - cmp_array[ERL_NEW_FUN_EXT] = 4; - cmp_array[ERL_PORT_EXT] = 5; - cmp_array[ERL_PID_EXT] = 6; - cmp_array[ERL_SMALL_TUPLE_EXT] = 7; - cmp_array[ERL_LARGE_TUPLE_EXT] = 7; - cmp_array[ERL_NIL_EXT] = 8; - cmp_array[ERL_STRING_EXT] = 9; - cmp_array[ERL_LIST_EXT] = 9; - cmp_array[ERL_BINARY_EXT] = 10; + cmp_array[ERL_SMALL_INTEGER_EXT] = ERL_NUM_CMP; + cmp_array[ERL_INTEGER_EXT] = ERL_NUM_CMP; + cmp_array[ERL_FLOAT_EXT] = ERL_NUM_CMP; + cmp_array[NEW_FLOAT_EXT] = ERL_NUM_CMP; + cmp_array[ERL_SMALL_BIG_EXT] = ERL_NUM_CMP; + cmp_array[ERL_LARGE_BIG_EXT] = ERL_NUM_CMP; + cmp_array[ERL_ATOM_EXT] = ERL_ATOM_CMP; + cmp_array[ERL_ATOM_UTF8_EXT] = ERL_ATOM_CMP; + cmp_array[ERL_SMALL_ATOM_EXT] = ERL_ATOM_CMP; + cmp_array[ERL_SMALL_ATOM_UTF8_EXT] = ERL_ATOM_CMP; + cmp_array[ERL_REFERENCE_EXT] = ERL_REF_CMP; + cmp_array[ERL_NEW_REFERENCE_EXT] = ERL_REF_CMP; + cmp_array[ERL_FUN_EXT] = ERL_FUN_CMP; + cmp_array[ERL_NEW_FUN_EXT] = ERL_FUN_CMP; + cmp_array[ERL_PORT_EXT] = ERL_PORT_CMP; + cmp_array[ERL_PID_EXT] = ERL_PID_CMP; + cmp_array[ERL_SMALL_TUPLE_EXT] = ERL_TUPLE_CMP; + cmp_array[ERL_LARGE_TUPLE_EXT] = ERL_TUPLE_CMP; + cmp_array[ERL_NIL_EXT] = ERL_NIL_CMP; + cmp_array[ERL_STRING_EXT] = ERL_LIST_CMP; + cmp_array[ERL_LIST_EXT] = ERL_LIST_CMP; + cmp_array[ERL_BINARY_EXT] = ERL_BIN_CMP; init_cmp_array_p = 0; } if (init_cmp_num_class_p) { @@ -156,6 +166,21 @@ static int erl_length_x(const ETERM *ep) { *============================================================== */ +static void encode_atom(Erl_Atom_data* a, unsigned char **ext) +{ + int ix = 0; + if (a->latin1) { + ei_encode_atom_len_as((char*)*ext, &ix, a->latin1, a->lenL, + ERLANG_LATIN1, ERLANG_LATIN1); + } + else if (ei_encode_atom_len_as((char*)*ext, &ix, a->utf8, a->lenU, + ERLANG_UTF8, ERLANG_LATIN1) < 0) { + ei_encode_atom_len_as((char*)*ext, &ix, a->utf8, a->lenU, + ERLANG_UTF8, ERLANG_UTF8); + } + *ext += ix; +} + /* * The actual ENCODE engine. * Returns 0 on success, otherwise 1. @@ -170,12 +195,7 @@ int erl_encode_it(ETERM *ep, unsigned char **ext, int dist) switch(ERL_TYPE(ep)) { case ERL_ATOM: - i = ep->uval.aval.len; - *(*ext)++ = ERL_ATOM_EXT; - *(*ext)++ = (i >>8) &0xff; - *(*ext)++ = i &0xff; - memcpy((void *) *ext, (const void *) ep->uval.aval.a, i); - *ext += i; + encode_atom(&ep->uval.aval.d, ext); return 0; case ERL_INTEGER: @@ -286,12 +306,7 @@ int erl_encode_it(ETERM *ep, unsigned char **ext, int dist) case ERL_PID: *(*ext)++ = ERL_PID_EXT; /* First poke in node as an atom */ - i = strlen((char *)ERL_PID_NODE(ep)); - *(*ext)++ = ERL_ATOM_EXT; - *(*ext)++ = (i >>8) &0xff; - *(*ext)++ = i &0xff; - memcpy(*ext, ERL_PID_NODE(ep), i); - *ext += i; + encode_atom(&ep->uval.pidval.node, ext); /* And then fill in the integer fields */ i = ERL_PID_NUMBER(ep); *(*ext)++ = (i >> 24) &0xff; @@ -319,11 +334,8 @@ int erl_encode_it(ETERM *ep, unsigned char **ext, int dist) *(*ext)++ = (len >> 8) &0xff; *(*ext)++ = len &0xff; - *(*ext)++ = ERL_ATOM_EXT; - *(*ext)++ = (i >> 8) &0xff; - *(*ext)++ = i &0xff; - memcpy(*ext, ERL_REF_NODE(ep), i); - *ext += i; + encode_atom(&ep->uval.refval.node, ext); + *(*ext)++ = ERL_REF_CREATION(ep); /* Then the integer fields */ for (j = 0; j < ERL_REF_LEN(ep); j++) { @@ -338,12 +350,7 @@ int erl_encode_it(ETERM *ep, unsigned char **ext, int dist) case ERL_PORT: *(*ext)++ = ERL_PORT_EXT; /* First poke in node as an atom */ - i = strlen((char *)ERL_PORT_NODE(ep)); - *(*ext)++ = ERL_ATOM_EXT; - *(*ext)++ = (i >>8) &0xff; - *(*ext)++ = i &0xff; - memcpy(*ext, ERL_PORT_NODE(ep), i); - *ext += i; + encode_atom(&ep->uval.portval.node, ext); /* Then the integer fields */ i = ERL_PORT_NUMBER(ep); *(*ext)++ = (i >> 24) &0xff; @@ -494,6 +501,16 @@ int erl_term_len(ETERM *ep) return 1+erl_term_len_helper(ep, 4); } +static int atom_len_helper(Erl_Atom_data* a) +{ + if (erl_atom_ptr_latin1(a)) { + return 1 + 2 + a->lenL; /* ERL_ATOM_EXT */ + } + else { + return 1 + 1 + (a->lenU > 255) + a->lenU; + } +} + static int erl_term_len_helper(ETERM *ep, int dist) { int len = 0; @@ -505,8 +522,7 @@ static int erl_term_len_helper(ETERM *ep, int dist) if (ep) { switch (ERL_TYPE(ep)) { case ERL_ATOM: - i = ep->uval.aval.len; - len = i + 3; + len = atom_len_helper(&ep->uval.aval.d); break; case ERL_INTEGER: @@ -538,20 +554,15 @@ static int erl_term_len_helper(ETERM *ep, int dist) break; case ERL_PID: - /* 1 + N + 4 + 4 + 1 where N = 3 + strlen */ - i = strlen((char *)ERL_PID_NODE(ep)); - len = 13 + i; + len = 1 + atom_len_helper(&ep->uval.pidval.node) + 4 + 4 + 1; break; case ERL_REF: - i = strlen((char *)ERL_REF_NODE(ep)); - len = 1 + 2 + (i+3) + 1 + ERL_REF_LEN(ep) * 4; + len = 1 + 2 + atom_len_helper(&ep->uval.refval.node) + 1 + ERL_REF_LEN(ep) * 4; break; case ERL_PORT: - /* 1 + N + 4 + 1 where N = 3 + strlen */ - i = strlen((char *)ERL_PORT_NODE(ep)); - len = 9 + i; + len = 1 + atom_len_helper(&ep->uval.portval.node) + 4 + 1; break; case ERL_EMPTY_LIST: @@ -644,31 +655,36 @@ int erl_encode_buf(ETERM *ep, unsigned char **ext) } /* erl_encode_buf */ -/* - * A nice macro to make it look cleaner in the - * cases of PID's,PORT's and REF's below. - * It reads the NODE name from a buffer. - */ -#define READ_THE_NODE(ext,cp,len,i) \ -/* eat first atom, repr. the node */ \ -if (**ext != ERL_ATOM_EXT) \ - return (ETERM *) NULL; \ -*ext += 1; \ -i = (**ext << 8) | (*ext)[1]; \ -cp = (char *) *(ext) + 2; \ -*ext += (i + 2); \ -len = i - -#define STATIC_NODE_BUF_SZ 30 - -#define SET_NODE(node,node_buf,cp,len) \ -if (len >= STATIC_NODE_BUF_SZ) node = erl_malloc(len+1); \ -else node = node_buf; \ -memcpy(node, cp, len); \ -node[len] = '\0' - -#define RESET_NODE(node,len) \ -if (len >= STATIC_NODE_BUF_SZ) free(node) + +static int read_atom(unsigned char** ext, Erl_Atom_data* a) +{ + char buf[MAXATOMLEN_UTF8]; + int offs = 0; + erlang_char_encoding enc; + int ret = ei_decode_atom_as((char*)*ext, &offs, buf, MAXATOMLEN_UTF8, + ERLANG_LATIN1|ERLANG_UTF8, NULL, &enc); + *ext += offs; + + if (ret == 0) { + int i = strlen(buf); + char* clone = erl_malloc(i+1); + memcpy(clone, buf, i+1); + + a->latin1 = NULL; + a->lenL = 0; + a->utf8 = NULL; + a->lenU = 0; + if (enc & (ERLANG_LATIN1 | ERLANG_ASCII)) { + a->latin1 = clone; + a->lenL = i; + } + if (enc & (ERLANG_UTF8 | ERLANG_ASCII)) { + a->utf8 = clone; + a->lenU = i; + } + } + return ret; +} /* * The actual DECODE engine. @@ -679,13 +695,13 @@ static ETERM *erl_decode_it(unsigned char **ext) char *cp; ETERM *ep,*tp,*np; unsigned int u,sign; - int i,j,len,arity; + int i,j,arity; double ff; /* Assume we are going to decode an integer */ ep = erl_alloc_eterm(ERL_INTEGER); ERL_COUNT(ep) = 1; - + switch (*(*ext)++) { case ERL_INTEGER_EXT: @@ -774,138 +790,90 @@ static ETERM *erl_decode_it(unsigned char **ext) return ep; case ERL_ATOM_EXT: + case ERL_SMALL_ATOM_EXT: + case ERL_ATOM_UTF8_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: + ERL_TYPE(ep) = ERL_ATOM; - i = (**ext << 8) | (*ext)[1]; - cp = (char *) *(ext) + 2; - *ext += (i + 2); - ep->uval.aval.len = i; - ep->uval.aval.a = (char *) erl_malloc(i+1); - memcpy(ep->uval.aval.a, cp, i); - ep->uval.aval.a[i]='\0'; + --(*ext); + if (read_atom(ext, &ep->uval.aval.d) < 0) return NULL; return ep; case ERL_PID_EXT: - erl_free_term(ep); - { /* Why not use the constructors? */ - char *node; - char node_buf[STATIC_NODE_BUF_SZ]; + { unsigned int number, serial; unsigned char creation; - ETERM *eterm_p; - READ_THE_NODE(ext,cp,len,i); - SET_NODE(node,node_buf,cp,len); + ERL_TYPE(ep) = ERL_PID; + if (read_atom(ext, &ep->uval.pidval.node) < 0) return NULL; /* get the integers */ -#if 0 - /* FIXME: Remove code or whatever.... - Ints on the wire are big-endian (== network byte order) - so use ntoh[sl]. (But some are little-endian! Arrrgh!) - Also, the libc authors can be expected to optimize them - heavily. However, the marshalling makes no guarantees - about alignments -- so it won't work at all. */ - number = ntohl(*((unsigned int *)*ext)++); - serial = ntohl(*((unsigned int *)*ext)++); -#else number = ((*ext)[0] << 24) | ((*ext)[1]) << 16 | ((*ext)[2]) << 8 | ((*ext)[3]); *ext += 4; serial = ((*ext)[0] << 24) | ((*ext)[1]) << 16 | ((*ext)[2]) << 8 | ((*ext)[3]); *ext += 4; -#endif creation = *(*ext)++; - eterm_p = erl_mk_pid(node, number, serial, creation); - RESET_NODE(node,len); - return eterm_p; + erl_mk_pid_helper(ep, number, serial, creation); + return ep; } case ERL_REFERENCE_EXT: - erl_free_term(ep); { - char *node; - char node_buf[STATIC_NODE_BUF_SZ]; - unsigned int number; + unsigned int n[3] = {0, 0, 0}; unsigned char creation; - ETERM *eterm_p; - READ_THE_NODE(ext,cp,len,i); - SET_NODE(node,node_buf,cp,len); + ERL_TYPE(ep) = ERL_REF; + if (read_atom(ext, &ep->uval.refval.node) < 0) return NULL; /* get the integers */ -#if 0 - number = ntohl(*((unsigned int *)*ext)++); -#else - number = ((*ext)[0] << 24) | ((*ext)[1]) << 16 | + n[0] = ((*ext)[0] << 24) | ((*ext)[1]) << 16 | ((*ext)[2]) << 8 | ((*ext)[3]); *ext += 4; -#endif creation = *(*ext)++; - eterm_p = erl_mk_ref(node, number, creation); - RESET_NODE(node,len); - return eterm_p; + __erl_mk_reference(ep, NULL, 1, n, creation); + return ep; } case ERL_NEW_REFERENCE_EXT: - erl_free_term(ep); { - char *node; - char node_buf[STATIC_NODE_BUF_SZ]; size_t cnt, i; unsigned int n[3]; unsigned char creation; - ETERM *eterm_p; -#if 0 - cnt = ntohs(*((unsigned short *)*ext)++); -#else + ERL_TYPE(ep) = ERL_REF; cnt = ((*ext)[0] << 8) | (*ext)[1]; *ext += 2; -#endif - READ_THE_NODE(ext,cp,len,i); - SET_NODE(node,node_buf,cp,len); + if (read_atom(ext, &ep->uval.refval.node) < 0) return NULL; /* get the integers */ creation = *(*ext)++; for(i = 0; i < cnt; i++) { -#if 0 - n[i] = ntohl(*((unsigned int *)*ext)++); -#else n[i] = ((*ext)[0] << 24) | ((*ext)[1]) << 16 | ((*ext)[2]) << 8 | ((*ext)[3]); *ext += 4; -#endif } - eterm_p = __erl_mk_reference(node, cnt, n, creation); - RESET_NODE(node,len); - return eterm_p; + __erl_mk_reference(ep, NULL, cnt, n, creation); + return ep; } case ERL_PORT_EXT: - erl_free_term(ep); { - char *node; - char node_buf[STATIC_NODE_BUF_SZ]; unsigned int number; unsigned char creation; - ETERM *eterm_p; - READ_THE_NODE(ext,cp,len,i); - SET_NODE(node,node_buf,cp,len); + ERL_TYPE(ep) = ERL_PORT; + if (read_atom(ext, &ep->uval.portval.node) < 0) return NULL; /* get the integers */ -#if 0 - number = ntohl(*((unsigned int *)*ext)++); -#else number = ((*ext)[0] << 24) | ((*ext)[1]) << 16 | ((*ext)[2]) << 8 | ((*ext)[3]); *ext += 4; -#endif creation = *(*ext)++; - eterm_p = erl_mk_port(node, number, creation); - RESET_NODE(node,len); - return eterm_p; + erl_mk_port_helper(ep, number, creation); + return ep; } case ERL_NIL_EXT: @@ -1140,6 +1108,9 @@ unsigned char erl_ext_type(unsigned char *ext) case ERL_INTEGER_EXT: return ERL_INTEGER; case ERL_ATOM_EXT: + case ERL_ATOM_UTF8_EXT: + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: return ERL_ATOM; case ERL_PID_EXT: return ERL_PID; @@ -1191,6 +1162,9 @@ int erl_ext_size(unsigned char *t) case ERL_SMALL_INTEGER_EXT: case ERL_INTEGER_EXT: case ERL_ATOM_EXT: + case ERL_ATOM_UTF8_EXT: + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: case ERL_PID_EXT: case ERL_PORT_EXT: case ERL_REFERENCE_EXT: @@ -1229,15 +1203,32 @@ int erl_ext_size(unsigned char *t) } /* ext_size */ -/* - * A nice macro that eats up the atom pointed to. - */ -#define JUMP_ATOM(ext,i) \ -if (**ext != ERL_ATOM_EXT) \ - return 0; \ -*ext += 1; \ -i = (**ext << 8) | (*ext)[1]; \ -*ext += (i + 2) + +static int jump_atom(unsigned char** ext) +{ + unsigned char* e = *ext; + int len; + + switch (*e++) { + case ERL_ATOM_EXT: + case ERL_ATOM_UTF8_EXT: + len = (e[0] << 8) | e[1]; + e += (len + 2); + break; + + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: + len = e[0]; + e += (len + 1); + break; + + default: + return 0; + } + *ext = e; + return 1; +} + /* * MOVE the POINTER PAST the ENCODED ETERM we @@ -1259,25 +1250,27 @@ static int jump(unsigned char **ext) *ext += 1; break; case ERL_ATOM_EXT: - i = (**ext << 8) | (*ext)[1]; - *ext += (i + 2); + case ERL_ATOM_UTF8_EXT: + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: + jump_atom(ext); break; case ERL_PID_EXT: /* eat first atom */ - JUMP_ATOM(ext,i); + if (!jump_atom(ext)) return 0; *ext += 9; /* Two int's and the creation field */ break; case ERL_REFERENCE_EXT: case ERL_PORT_EXT: /* first field is an atom */ - JUMP_ATOM(ext,i); + if (!jump_atom(ext)) return 0; *ext += 5; /* One int and the creation field */ break; case ERL_NEW_REFERENCE_EXT: n = (**ext << 8) | (*ext)[1]; *ext += 2; /* first field is an atom */ - JUMP_ATOM(ext,i); + if (!jump_atom(ext)) return 0; *ext += 4*n+1; break; case ERL_NIL_EXT: @@ -1425,6 +1418,58 @@ static int cmpbytes(unsigned char* s1,int l1,unsigned char* s2,int l2) } /* cmpbytes */ +#define tag2enc(T) ((T)==ERL_ATOM_EXT || (T)==ERL_SMALL_ATOM_EXT ? ERLANG_LATIN1 : ERLANG_UTF8) + +static int cmpatoms(unsigned char* s1, int l1, unsigned char tag1, + unsigned char* s2, int l2, unsigned char tag2) +{ + erlang_char_encoding enc1 = tag2enc(tag1); + erlang_char_encoding enc2 = tag2enc(tag2); + + if (enc1 == enc2) { + return cmpbytes(s1, l1,s2,l2); + } + + if (enc1 == ERLANG_LATIN1) { + return cmp_latin1_vs_utf8((char*)s1, l1, (char*)s2, l2); + } + else { + return -cmp_latin1_vs_utf8((char*)s2, l2, (char*)s1, l1); + } +} + +int cmp_latin1_vs_utf8(const char* strL, int lenL, const char* strU, int lenU) +{ + unsigned char* sL = (unsigned char*)strL; + unsigned char* sU = (unsigned char*)strU; + unsigned char* sL_end = sL + lenL; + unsigned char* sU_end = sU + lenU; + + while(sL < sL_end && sU < sU_end) { + unsigned char UasL; + if (*sL >= 0x80) { + if (*sU < 0xC4 && (sU+1) < sU_end) { + UasL = ((sU[0] & 0x3) << 6) | (sU[1] & 0x3F); + } + else return -1; + } + else { + UasL = *sU; + } + if (*sL < UasL) return -1; + if (*sL > UasL) return 1; + + sL++; + if (*sU < 0x80) sU++; + else if (*sU < 0xE0) sU += 2; + else if (*sU < 0xF0) sU += 3; + else /*if (*sU < 0xF8)*/ sU += 4; + } + + return (sU >= sU_end) - (sL >= sL_end); /* -1, 0 or 1 */ +} + + #define CMP_EXT_ERROR_CODE 4711 #define CMP_EXT_INT32_BE(AP, BP) \ @@ -1437,9 +1482,8 @@ do { \ #define CMP_EXT_SKIP_ATOM(EP) \ do { \ - if ((EP)[0] != ERL_ATOM_EXT) \ + if (!jump_atom(&(EP))) \ return CMP_EXT_ERROR_CODE; \ - (EP) += 3 + ((EP)[1] << 8 | (EP)[2]); \ } while (0) /* @@ -1561,6 +1605,7 @@ static int cmp_exe2(unsigned char **e1, unsigned char **e2) int min, ret,i,j,k; double ff1, ff2; unsigned char *tmp1, *tmp2; + unsigned char tag1, tag2; if ( ((*e1)[0] == ERL_STRING_EXT) && ((*e2)[0] == ERL_LIST_EXT) ) { return cmp_string_list(e1, e2); @@ -1568,8 +1613,10 @@ static int cmp_exe2(unsigned char **e1, unsigned char **e2) return -cmp_string_list(e2, e1); } - *e2 += 1; - switch (*(*e1)++) + tag1 = *(*e1)++; + tag2 = *(*e2)++; + i = j = 0; + switch (tag1) { case ERL_SMALL_INTEGER_EXT: if (**e1 < **e2) ret = -1; @@ -1589,11 +1636,17 @@ static int cmp_exe2(unsigned char **e1, unsigned char **e2) *e1 += 4; *e2 += 4; return ret; case ERL_ATOM_EXT: - i = (**e1 << 8) | (*e1)[1]; - j = (**e2 << 8) | (*e2)[1]; - ret = cmpbytes(*e1 +2, i, *e2 +2, j); - *e1 += (i + 2); - *e2 += (j + 2); + case ERL_ATOM_UTF8_EXT: + i = (**e1) << 8; (*e1)++; + j = (**e2) << 8; (*e2)++; + /*fall through*/ + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: + i |= (**e1); (*e1)++; + j |= (**e2); (*e2)++; + ret = cmpatoms(*e1, i, tag1, *e2, j, tag2); + *e1 += i; + *e2 += j; return ret; case ERL_PID_EXT: { unsigned char *n1 = *e1; @@ -1622,7 +1675,7 @@ static int cmp_exe2(unsigned char **e1, unsigned char **e2) } case ERL_PORT_EXT: /* First compare node names ... */ - if (**e1 != ERL_ATOM_EXT || **e2 != ERL_ATOM_EXT) + if (!IS_ERL_ATOM(**e1) || !IS_ERL_ATOM(**e2)) return CMP_EXT_ERROR_CODE; ret = cmp_exe2(e1, e2); *e1 += 5; *e2 += 5; diff --git a/lib/erl_interface/src/legacy/global_whereis.c b/lib/erl_interface/src/legacy/global_whereis.c index 2afb193504..589c9d17de 100644 --- a/lib/erl_interface/src/legacy/global_whereis.c +++ b/lib/erl_interface/src/legacy/global_whereis.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2009. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -85,7 +85,16 @@ ETERM *erl_global_whereis(int fd, const char *name, char *node) opid = erl_decode((unsigned char*)buf); /* extract the nodename for the caller */ - if (node) strcpy(node,epid.node); + if (node) { + char* node_str = ERL_PID_NODE(opid); + if (node_str) { + strcpy(node, node_str); + } + else { + erl_free_term(opid); + return NULL; + } + } return opid; } diff --git a/lib/erl_interface/src/misc/ei_decode_term.c b/lib/erl_interface/src/misc/ei_decode_term.c index 0b82ef0e35..1423ec7ed7 100644 --- a/lib/erl_interface/src/misc/ei_decode_term.c +++ b/lib/erl_interface/src/misc/ei_decode_term.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2011. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -32,7 +32,7 @@ int ei_decode_ei_term(const char* buf, int* index, ei_term* term) { const char* s = buf + *index, * s0 = s; - int len, i, n, sign; + int i, n, sign; char c; if (term == NULL) return -1; @@ -48,20 +48,13 @@ int ei_decode_ei_term(const char* buf, int* index, ei_term* term) case NEW_FLOAT_EXT: return ei_decode_double(buf, index, &term->value.d_val); case ERL_ATOM_EXT: - len = get16be(s); - if (len > MAXATOMLEN) return -1; - memcpy(term->value.atom_name, s, len); - term->value.atom_name[len] = '\0'; - s += len; - break; + case ERL_ATOM_UTF8_EXT: + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: + return ei_decode_atom(buf, index, term->value.atom_name); case ERL_REFERENCE_EXT: /* first the nodename */ - if (get8(s) != ERL_ATOM_EXT) return -1; - len = get16be(s); - if (len > MAXATOMLEN) return -1; - memcpy(term->value.ref.node, s, len); - term->value.ref.node[len] = '\0'; - s += len; + if (get_atom(&s, term->value.ref.node, &term->value.ref.node_org_enc) < 0) return -1; /* now the numbers: num (4), creation (1) */ term->value.ref.n[0] = get32be(s); term->value.ref.len = 1; @@ -71,12 +64,7 @@ int ei_decode_ei_term(const char* buf, int* index, ei_term* term) /* first the integer count */ term->value.ref.len = get16be(s); /* then the nodename */ - if (get8(s) != ERL_ATOM_EXT) return -1; - len = get16be(s); - if (len > MAXATOMLEN) return -1; - memcpy(term->value.ref.node, s, len); - term->value.ref.node[len] = '\0'; - s += len; + if (get_atom(&s, term->value.ref.node, &term->value.ref.node_org_enc) < 0) return -1; /* creation */ term->value.ref.creation = get8(s) & 0x03; /* finally the id integers */ @@ -88,22 +76,12 @@ int ei_decode_ei_term(const char* buf, int* index, ei_term* term) } break; case ERL_PORT_EXT: - if (get8(s) != ERL_ATOM_EXT) return -1; - len = get16be(s); - if (len > MAXATOMLEN) return -1; - memcpy(term->value.port.node, s, len); - term->value.port.node[len] = '\0'; + if (get_atom(&s, term->value.port.node, &term->value.port.node_org_enc) < 0) return -1; term->value.port.id = get32be(s) & 0x0fffffff; /* 28 bits */; term->value.port.creation = get8(s) & 0x03; break; case ERL_PID_EXT: - if (get8(s) != ERL_ATOM_EXT) return -1; - /* name first */ - len = get16be(s); - if (len > MAXATOMLEN) return -1; - memcpy(term->value.pid.node, s, len); - term->value.pid.node[len] = '\0'; - s += len; + if (get_atom(&s, term->value.pid.node, &term->value.port.node_org_enc) < 0) return -1; /* now the numbers: num (4), serial (4), creation (1) */ term->value.pid.num = get32be(s) & 0x7fff; /* 15 bits */ term->value.pid.serial = get32be(s) & 0x1fff; /* 13 bits */ diff --git a/lib/erl_interface/src/misc/ei_format.c b/lib/erl_interface/src/misc/ei_format.c index 281a192535..07499213cd 100644 --- a/lib/erl_interface/src/misc/ei_format.c +++ b/lib/erl_interface/src/misc/ei_format.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2012. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -139,8 +139,8 @@ static int patom(const char** fmt, ei_x_buff* x) --(*fmt); len = *fmt - start; /* FIXME why truncate atom name and not fail?! */ - if (len > MAXATOMLEN) - len = MAXATOMLEN; + if (len >= MAXATOMLEN) + len = MAXATOMLEN-1; return ei_x_encode_atom_len(x, start, len); } diff --git a/lib/erl_interface/src/misc/ei_printterm.c b/lib/erl_interface/src/misc/ei_printterm.c index 5fc6b3542c..95369db15a 100644 --- a/lib/erl_interface/src/misc/ei_printterm.c +++ b/lib/erl_interface/src/misc/ei_printterm.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2010. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -115,7 +115,7 @@ static int print_term(FILE* fp, ei_x_buff* x, const char* buf, int* index) { int i, doquote, n, m, ty, r; - char a[MAXATOMLEN+1], *p; + char a[MAXATOMLEN], *p; int ch_written = 0; /* counter of written chars */ erlang_pid pid; erlang_port port; @@ -132,7 +132,10 @@ static int print_term(FILE* fp, ei_x_buff* x, doquote = 0; ei_get_type_internal(buf, index, &ty, &n); switch (ty) { - case ERL_ATOM_EXT: + case ERL_ATOM_EXT: + case ERL_ATOM_UTF8_EXT: + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: if (ei_decode_atom(buf, index, a) < 0) goto err; doquote = !islower((int)a[0]); diff --git a/lib/erl_interface/src/misc/ei_x_encode.c b/lib/erl_interface/src/misc/ei_x_encode.c index fa1e26ccbb..14d0b56b8f 100644 --- a/lib/erl_interface/src/misc/ei_x_encode.c +++ b/lib/erl_interface/src/misc/ei_x_encode.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2001-2009. All Rights Reserved. + * Copyright Ericsson AB 2001-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -99,7 +99,8 @@ int ei_x_encode_string(ei_x_buff* x, const char* s) int ei_x_encode_string_len(ei_x_buff* x, const char* s, int len) { int i = x->index; - ei_encode_string_len(NULL, &i, s, len); + if (ei_encode_string_len(NULL, &i, s, len) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_string_len(x->buff, &x->index, s, len); @@ -108,7 +109,8 @@ int ei_x_encode_string_len(ei_x_buff* x, const char* s, int len) int ei_x_encode_binary(ei_x_buff* x, const void* p, int len) { int i = x->index; - ei_encode_binary(NULL, &i, p, len); + if (ei_encode_binary(NULL, &i, p, len) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_binary(x->buff, &x->index, p, len); @@ -117,7 +119,8 @@ int ei_x_encode_binary(ei_x_buff* x, const void* p, int len) int ei_x_encode_long(ei_x_buff* x, long n) { int i = x->index; - ei_encode_long(NULL, &i, n); + if (ei_encode_long(NULL, &i, n) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_long(x->buff, &x->index, n); @@ -126,7 +129,8 @@ int ei_x_encode_long(ei_x_buff* x, long n) int ei_x_encode_ulong(ei_x_buff* x, unsigned long n) { int i = x->index; - ei_encode_ulong(NULL, &i, n); + if (ei_encode_ulong(NULL, &i, n) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_ulong(x->buff, &x->index, n); @@ -135,7 +139,8 @@ int ei_x_encode_ulong(ei_x_buff* x, unsigned long n) int ei_x_encode_char(ei_x_buff* x, char p) { int i = x->index; - ei_encode_char(NULL, &i, p); + if (ei_encode_char(NULL, &i, p) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_char(x->buff, &x->index, p); @@ -144,7 +149,8 @@ int ei_x_encode_char(ei_x_buff* x, char p) int ei_x_encode_boolean(ei_x_buff* x, int p) { int i = x->index; - ei_encode_boolean(NULL, &i, p); + if (ei_encode_boolean(NULL, &i, p) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_boolean(x->buff, &x->index, p); @@ -153,7 +159,8 @@ int ei_x_encode_boolean(ei_x_buff* x, int p) int ei_x_encode_double(ei_x_buff* x, double dbl) { int i = x->index; - ei_encode_double(NULL, &i, dbl); + if (ei_encode_double(NULL, &i, dbl) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_double(x->buff, &x->index, dbl); @@ -162,7 +169,8 @@ int ei_x_encode_double(ei_x_buff* x, double dbl) int ei_x_encode_list_header(ei_x_buff* x, long n) { int i = x->index; - ei_encode_list_header(NULL, &i, n); + if (ei_encode_list_header(NULL, &i, n) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_list_header(x->buff, &x->index, n); @@ -171,7 +179,8 @@ int ei_x_encode_list_header(ei_x_buff* x, long n) int ei_x_encode_empty_list(ei_x_buff* x) { int i = x->index; - ei_encode_empty_list(NULL, &i); + if (ei_encode_empty_list(NULL, &i) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_empty_list(x->buff, &x->index); @@ -180,7 +189,8 @@ int ei_x_encode_empty_list(ei_x_buff* x) int ei_x_encode_version(ei_x_buff* x) { int i = x->index; - ei_encode_version(NULL, &i); + if (ei_encode_version(NULL, &i) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_version(x->buff, &x->index); @@ -189,7 +199,8 @@ int ei_x_encode_version(ei_x_buff* x) int ei_x_encode_tuple_header(ei_x_buff* x, long n) { int i = x->index; - ei_encode_tuple_header(NULL, &i, n); + if (ei_encode_tuple_header(NULL, &i, n) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_tuple_header(x->buff, &x->index, n); @@ -197,22 +208,39 @@ int ei_x_encode_tuple_header(ei_x_buff* x, long n) int ei_x_encode_atom(ei_x_buff* x, const char* s) { - return ei_x_encode_atom_len(x, s, strlen(s)); + return ei_x_encode_atom_len_as(x, s, strlen(s), ERLANG_LATIN1, ERLANG_LATIN1); } int ei_x_encode_atom_len(ei_x_buff* x, const char* s, int len) { + return ei_x_encode_atom_len_as(x, s, len, ERLANG_LATIN1, ERLANG_LATIN1); +} + +int ei_x_encode_atom_as(ei_x_buff* x, const char* s, + erlang_char_encoding from_enc, + erlang_char_encoding to_enc) +{ + return ei_x_encode_atom_len_as(x, s, strlen(s), from_enc, to_enc); +} + +int ei_x_encode_atom_len_as(ei_x_buff* x, const char* s, int len, + erlang_char_encoding from_enc, + erlang_char_encoding to_enc) +{ int i = x->index; - ei_encode_atom_len(NULL, &i, s, len); + if (ei_encode_atom_len_as(NULL, &i, s, len, from_enc, to_enc) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; - return ei_encode_atom_len(x->buff, &x->index, s, len); + return ei_encode_atom_len_as(x->buff, &x->index, s, len, from_enc, to_enc); } + int ei_x_encode_pid(ei_x_buff* x, const erlang_pid* pid) { int i = x->index; - ei_encode_pid(NULL, &i, pid); + if (ei_encode_pid(NULL, &i, pid) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_pid(x->buff, &x->index, pid); @@ -221,7 +249,8 @@ int ei_x_encode_pid(ei_x_buff* x, const erlang_pid* pid) int ei_x_encode_fun(ei_x_buff* x, const erlang_fun* fun) { int i = x->index; - ei_encode_fun(NULL, &i, fun); + if (ei_encode_fun(NULL, &i, fun) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_fun(x->buff, &x->index, fun); @@ -230,7 +259,8 @@ int ei_x_encode_fun(ei_x_buff* x, const erlang_fun* fun) int ei_x_encode_ref(ei_x_buff* x, const erlang_ref* ref) { int i = x->index; - ei_encode_ref(NULL, &i, ref); + if (ei_encode_ref(NULL, &i, ref) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_ref(x->buff, &x->index, ref); @@ -239,7 +269,8 @@ int ei_x_encode_ref(ei_x_buff* x, const erlang_ref* ref) int ei_x_encode_port(ei_x_buff* x, const erlang_port* port) { int i = x->index; - ei_encode_port(NULL, &i, port); + if (ei_encode_port(NULL, &i, port) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_port(x->buff, &x->index, port); @@ -248,7 +279,8 @@ int ei_x_encode_port(ei_x_buff* x, const erlang_port* port) int ei_x_encode_trace(ei_x_buff* x, const erlang_trace* trace) { int i = x->index; - ei_encode_trace(NULL, &i, trace); + if (ei_encode_trace(NULL, &i, trace) == -1) + return -1; if (!x_fix_buff(x, i)) return -1; return ei_encode_trace(x->buff, &x->index, trace); diff --git a/lib/erl_interface/src/misc/get_type.c b/lib/erl_interface/src/misc/get_type.c index 2a680d0f94..2bd992bd56 100644 --- a/lib/erl_interface/src/misc/get_type.c +++ b/lib/erl_interface/src/misc/get_type.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2010. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -33,78 +33,6 @@ int ei_get_type(const char *buf, const int *index, int *type, int *len) return ei_get_type_internal(buf, index, type, len); } -#if 0 -int ei_get_type(const char *buf, const int *index, int *type, int *len) -{ - const char *s = buf + *index; - int itype = get8(s); /* Internal type */ - - *len = 0; - - switch (*type) { - - case ERL_SMALL_INTEGER_EXT: - case ERL_INTEGER_EXT: - case ERL_SMALL_BIG_EXT: - case ERL_LARGE_BIG_EXT: - *type = EI_TYPE_INTEGER; - break; - - case ERL_FLOAT_EXT: - *type = EI_TYPE_FLOAT; - break; - - case ERL_SMALL_TUPLE_EXT: - *len = get8(s); - break; - - case ERL_ATOM_EXT: - case ERL_STRING_EXT: - *len = get16be(s); - break; - - case ERL_LARGE_TUPLE_EXT: - case ERL_LIST_EXT: - case ERL_BINARY_EXT: - *len = get32be(s); - break; - - case ERL_SMALL_BIG_EXT: - *len = (get8(s)+1)/2; /* big arity */ - break; - - case ERL_LARGE_BIG_EXT: - *len = (get32be(s)+1)/2; /* big arity */ - break; - - case ERL_BINARY_EXT: - *type = EI_TYPE_BINARY; - break; - - case ERL_PID_EXT: - *type = EI_TYPE_PID; - break; - - case ERL_PORT_EXT: - *type = EI_TYPE_PORT; - break; - - case ERL_REFERENCE_EXT: - case ERL_NEW_REFERENCE_EXT: - *type = EI_TYPE_REF; - break; - - default: - break; - } - - /* leave index unchanged */ - return 0; -} -#endif - - -/* Old definition of function above */ int ei_get_type_internal(const char *buf, const int *index, int *type, int *len) @@ -114,10 +42,15 @@ int ei_get_type_internal(const char *buf, const int *index, *type = get8(s); switch (*type) { + case ERL_SMALL_ATOM_EXT: + case ERL_SMALL_ATOM_UTF8_EXT: + *type = ERL_ATOM_EXT; case ERL_SMALL_TUPLE_EXT: *len = get8(s); break; - + + case ERL_ATOM_UTF8_EXT: + *type = ERL_ATOM_EXT; case ERL_ATOM_EXT: case ERL_STRING_EXT: *len = get16be(s); diff --git a/lib/erl_interface/src/misc/putget.h b/lib/erl_interface/src/misc/putget.h index 7a43de324b..c751e03093 100644 --- a/lib/erl_interface/src/misc/putget.h +++ b/lib/erl_interface/src/misc/putget.h @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2010. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -105,6 +105,13 @@ ((EI_ULONGLONG)((unsigned char *)(s))[-2] << 8) | \ (EI_ULONGLONG)((unsigned char *)(s))[-1])) +int utf8_to_latin1(char* dst, const char* src, int slen, int destlen, erlang_char_encoding* res_encp); +int latin1_to_utf8(char* dst, const char* src, int slen, int destlen, erlang_char_encoding* res_encp); +int ei_internal_get_atom(const char** bufp, char* p, erlang_char_encoding*); +int ei_internal_put_atom(char** bufp, const char* p, int slen, erlang_char_encoding); +#define get_atom ei_internal_get_atom +#define put_atom ei_internal_put_atom + typedef union float_ext { double d; EI_ULONGLONG val; diff --git a/lib/erl_interface/src/misc/show_msg.c b/lib/erl_interface/src/misc/show_msg.c index 194296798b..935a9a3287 100644 --- a/lib/erl_interface/src/misc/show_msg.c +++ b/lib/erl_interface/src/misc/show_msg.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 1998-2010. All Rights Reserved. + * Copyright Ericsson AB 1998-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -132,13 +132,13 @@ int ei_show_sendmsg(FILE *stream, const char *header, const char *msgbuf) switch (msg.msgtype) { case ERL_SEND: - if (ei_decode_atom(header,&index,msg.cookie) + if (ei_decode_atom_as(header,&index,msg.cookie,sizeof(msg.cookie),ERLANG_UTF8,NULL,NULL) || ei_decode_pid(header,&index,&msg.to)) return -1; mbuf = msgbuf; break; case ERL_SEND_TT: - if (ei_decode_atom(header,&index,msg.cookie) + if (ei_decode_atom_as(header,&index,msg.cookie,sizeof(msg.cookie),ERLANG_UTF8,NULL,NULL) || ei_decode_pid(header,&index,&msg.to) || ei_decode_trace(header,&index,&msg.token)) return -1; mbuf = msgbuf; @@ -146,15 +146,15 @@ int ei_show_sendmsg(FILE *stream, const char *header, const char *msgbuf) case ERL_REG_SEND: if (ei_decode_pid(header,&index,&msg.from) - || ei_decode_atom(header,&index,msg.cookie) - || ei_decode_atom(header,&index,msg.toname)) return -1; + || ei_decode_atom_as(header,&index,msg.cookie,sizeof(msg.cookie),ERLANG_UTF8,NULL,NULL) + || ei_decode_atom_as(header,&index,msg.toname,sizeof(msg.toname),ERLANG_UTF8,NULL,NULL)) return -1; mbuf = msgbuf; break; case ERL_REG_SEND_TT: if (ei_decode_pid(header,&index,&msg.from) - || ei_decode_atom(header,&index,msg.cookie) - || ei_decode_atom(header,&index,msg.toname) + || ei_decode_atom_as(header,&index,msg.cookie,sizeof(msg.cookie),ERLANG_UTF8,NULL,NULL) + || ei_decode_atom_as(header,&index,msg.toname,sizeof(msg.toname),ERLANG_UTF8,NULL,NULL) || ei_decode_trace(header,&index,&msg.token)) return -1; mbuf = msgbuf; break; @@ -457,7 +457,7 @@ static void show_term(const char *termbuf, int *index, FILE *stream) break; case ERL_FUN_EXT: { - char atom[MAXATOMLEN+1]; + char atom[MAXATOMLEN]; long idx; long uniq; const char* s = termbuf + *index, * s0 = s; diff --git a/lib/erl_interface/src/prog/ei_fake_prog.c b/lib/erl_interface/src/prog/ei_fake_prog.c index 68eb537211..56d4eb7db4 100644 --- a/lib/erl_interface/src/prog/ei_fake_prog.c +++ b/lib/erl_interface/src/prog/ei_fake_prog.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2002-2009. All Rights Reserved. + * Copyright Ericsson AB 2002-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -96,6 +96,7 @@ int main(void) EI_ULONGLONG *ulonglongp = (EI_ULONGLONG*)NULL; EI_ULONGLONG ulonglongx = 0; #endif + erlang_char_encoding enc; intx = erl_errno; @@ -148,9 +149,13 @@ int main(void) ei_x_encode_string(&eix, charp); ei_x_encode_string_len(&eix, charp, intx); ei_encode_atom(charp, intp, charp); + ei_encode_atom_as(charp, intp, charp, ERLANG_LATIN1, ERLANG_UTF8); ei_encode_atom_len(charp, intp, charp, intx); + ei_encode_atom_len_as(charp, intp, charp, intx, ERLANG_ASCII, ERLANG_LATIN1); ei_x_encode_atom(&eix, charp); + ei_x_encode_atom_as(&eix, charp, ERLANG_LATIN1, ERLANG_UTF8); ei_x_encode_atom_len(&eix, charp, intx); + ei_x_encode_atom_len_as(&eix, charp, intx, ERLANG_LATIN1, ERLANG_UTF8); ei_encode_binary(charp, intp, (void *)0, longx); ei_x_encode_binary(&eix, (void*)0, intx); ei_encode_pid(charp, intp, &epid); @@ -181,6 +186,7 @@ int main(void) ei_decode_char(charp, intp, charp); ei_decode_string(charp, intp, charp); ei_decode_atom(charp, intp, charp); + ei_decode_atom_as(charp, intp, charp, MAXATOMLEN_UTF8, ERLANG_WHATEVER, &enc, &enc); ei_decode_binary(charp, intp, (void *)0, longp); ei_decode_fun(charp, intp, &efun); free_fun(&efun); diff --git a/lib/erl_interface/test/all_SUITE_data/init_tc.erl b/lib/erl_interface/test/all_SUITE_data/init_tc.erl index 7e370c566a..0f2ac215f9 100644 --- a/lib/erl_interface/test/all_SUITE_data/init_tc.erl +++ b/lib/erl_interface/test/all_SUITE_data/init_tc.erl @@ -40,7 +40,7 @@ run([]) -> run1(Name) -> CFile = Name ++ ".c", {ok, Bin} = file:read_file(CFile), - RE = "\nTESTCASE\\(([_a-zA-Z]*)\\)", + RE = "\nTESTCASE\\(([_a-zA-Z0-9]*)\\)", {match, Cases0} = re:run(Bin, RE, [{capture,all_but_first,list},global]), Cases = lists:concat(Cases0), generate(Name, Cases). diff --git a/lib/erl_interface/test/ei_decode_SUITE.erl b/lib/erl_interface/test/ei_decode_SUITE.erl index bb44b78854..5a94994912 100644 --- a/lib/erl_interface/test/ei_decode_SUITE.erl +++ b/lib/erl_interface/test/ei_decode_SUITE.erl @@ -1,7 +1,8 @@ +%% -*- coding: utf-8 -*- %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 2004-2011. All Rights Reserved. +%% Copyright Ericsson AB 2004-2012. All Rights Reserved. %% %% The contents of this file are subject to the Erlang Public License, %% Version 1.1, (the "License"); you may not use this file except in @@ -26,14 +27,16 @@ -export( [ all/0, suite/0,groups/0,init_per_suite/1, end_per_suite/1, - init_per_group/2,end_per_group/2, + init_per_group/2,end_per_group/2, init_per_testcase/2, + end_per_testcase/2, test_ei_decode_long/1, test_ei_decode_ulong/1, test_ei_decode_longlong/1, test_ei_decode_ulonglong/1, test_ei_decode_char/1, test_ei_decode_nonoptimal/1, - test_ei_decode_misc/1 + test_ei_decode_misc/1, + test_ei_decode_utf8_atom/1 ]). suite() -> [{ct_hooks,[ts_install_cth]}]. @@ -42,7 +45,7 @@ all() -> [test_ei_decode_long, test_ei_decode_ulong, test_ei_decode_longlong, test_ei_decode_ulonglong, test_ei_decode_char, test_ei_decode_nonoptimal, - test_ei_decode_misc]. + test_ei_decode_misc, test_ei_decode_utf8_atom]. groups() -> []. @@ -59,6 +62,11 @@ init_per_group(_GroupName, Config) -> end_per_group(_GroupName, Config) -> Config. +init_per_testcase(_TC, Config) -> + Config. + +end_per_testcase(_RC, Config) -> + Config. %% --------------------------------------------------------------------------- @@ -204,15 +212,15 @@ test_ei_decode_misc(Config) when is_list(Config) -> ?line send_term_as_binary(P,foo), ?line send_term_as_binary(P,''), - ?line send_term_as_binary(P,'������'), + ?line send_term_as_binary(P,'ÅÄÖåäö'), ?line send_term_as_binary(P,"foo"), ?line send_term_as_binary(P,""), - ?line send_term_as_binary(P,"������"), + ?line send_term_as_binary(P,"ÅÄÖåäö"), ?line send_term_as_binary(P,<<"foo">>), ?line send_term_as_binary(P,<<>>), - ?line send_term_as_binary(P,<<"������">>), + ?line send_term_as_binary(P,<<"ÅÄÖåäö">>), % ?line send_term_as_binary(P,{}), % ?line send_term_as_binary(P,[]), @@ -220,6 +228,29 @@ test_ei_decode_misc(Config) when is_list(Config) -> ?line runner:recv_eot(P), ok. +%% ######################################################################## %% + +test_ei_decode_utf8_atom(Config) -> + ?line P = runner:start(?test_ei_decode_utf8_atom), + + send_utf8_atom_as_binary(P,"å"), + send_utf8_atom_as_binary(P,"ä"), + send_term_as_binary(P,'ö'), + send_term_as_binary(P,'õ'), + + ?line send_utf8_atom_as_binary(P,[1758]), + ?line send_utf8_atom_as_binary(P,[1758,1758]), + ?line send_utf8_atom_as_binary(P,[1758,1758,1758]), + ?line send_utf8_atom_as_binary(P,[1758,1758,1758,1758]), + + send_utf8_atom_as_binary(P,"a"), + send_utf8_atom_as_binary(P,"b"), + send_term_as_binary(P,'c'), + send_term_as_binary(P,'d'), + + ?line runner:recv_eot(P), + ok. + %% ######################################################################## %% @@ -229,6 +260,8 @@ send_term_as_binary(Port, Term) when is_port(Port) -> send_raw(Port, Bin) when is_port(Port) -> Port ! {self(), {command, Bin}}. +send_utf8_atom_as_binary(Port, String) -> + Port ! {self(), {command, term_to_binary(uc_atup(String))}}. send_integers(P) -> ?line send_term_as_binary(P,0), % SMALL_INTEGER_EXT smallest @@ -303,3 +336,43 @@ send_integers2(P) -> ?line send_term_as_binary(P, 16#ffffffffffffffff), % largest u64 ?line send_term_as_binary(P, []), % illegal type ok. + +uc_atup(ATxt) -> + string_to_atom(ATxt). + +string_to_atom(String) -> + Utf8List = string_to_utf8_list(String), + Len = length(Utf8List), + TagLen = case Len < 256 of + true -> [119, Len]; + false -> [118, Len bsr 8, Len band 16#ff] + end, + binary_to_term(list_to_binary([131, TagLen, Utf8List])). + +string_to_utf8_list([]) -> + []; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 0 =< CP, + CP =< 16#7F -> + [CP | string_to_utf8_list(CPs)]; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 16#80 =< CP, + CP =< 16#7FF -> + [16#C0 bor (CP bsr 6), + 16#80 bor (16#3F band CP) + | string_to_utf8_list(CPs)]; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 16#800 =< CP, + CP =< 16#FFFF -> + [16#E0 bor (CP bsr 12), + 16#80 bor (16#3F band (CP bsr 6)), + 16#80 bor (16#3F band CP) + | string_to_utf8_list(CPs)]; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 16#10000 =< CP, + CP =< 16#10FFFF -> + [16#F0 bor (CP bsr 18), + 16#80 bor (16#3F band (CP bsr 12)), + 16#80 bor (16#3F band (CP bsr 6)), + 16#80 bor (16#3F band CP) + | string_to_utf8_list(CPs)]. diff --git a/lib/erl_interface/test/ei_decode_SUITE_data/ei_decode_test.c b/lib/erl_interface/test/ei_decode_SUITE_data/ei_decode_test.c index b349138ae9..f5c8c4fa7d 100644 --- a/lib/erl_interface/test/ei_decode_SUITE_data/ei_decode_test.c +++ b/lib/erl_interface/test/ei_decode_SUITE_data/ei_decode_test.c @@ -17,6 +17,8 @@ * %CopyrightEnd% */ +#include <string.h> + #ifdef VXWORKS #include "reclaim.h" #endif @@ -38,6 +40,24 @@ err, size1, SIZE, (EI_LONGLONG)p); #endif +#define ERLANG_ANY (ERLANG_ASCII|ERLANG_LATIN1|ERLANG_UTF8) + +struct my_atom { + erlang_char_encoding from; + erlang_char_encoding was_check; + erlang_char_encoding result_check; +}; + +/* Allow arrays constants to be part of macro arguments */ +#define P99(...) __VA_ARGS__ + +int ei_decode_my_atom_as(const char *buf, int *index, char *to, + struct my_atom *atom); +int ei_decode_my_atom(const char *buf, int *index, char *to, + struct my_atom *atom); +int ei_decode_my_string(const char *buf, int *index, char *to, + struct my_atom *atom); + #define EI_DECODE_2(FUNC,SIZE,TYPE,VAL) \ { \ TYPE p; \ @@ -129,17 +149,26 @@ } \ } \ -#define EI_DECODE_STRING(FUNC,SIZE,VAL) \ +#define dump(arr, num) { \ + int i; \ + message("Dumping " #arr ": "); \ + for (i = 0; i < num; i++) message("%u, ",(unsigned char)arr[i]); \ + message("\n"); \ + } + +#define EI_DECODE_STRING_4(FUNC,SIZE,VAL,ATOM) \ { \ char p[1024]; \ char *buf; \ + unsigned char val[] = VAL; \ int size1 = 0; \ int size2 = 0; \ int err; \ - message("ei_" #FUNC " should be " #VAL); \ + struct my_atom atom = ATOM; \ + message("ei_" #FUNC " should be " #VAL "\n"); \ buf = read_packet(NULL); \ \ - err = ei_ ## FUNC(buf+1, &size1, NULL); \ + err = ei_ ## FUNC(buf+1, &size1, NULL, &atom); \ message("err = %d, size = %d, expected size = %d\n",err,size1,SIZE); \ if (err != 0) { \ if (err != -1) { \ @@ -150,7 +179,7 @@ return; \ } \ \ - err = ei_ ## FUNC(buf+1, &size2, p); \ + err = ei_ ## FUNC(buf+1, &size2, p, &atom); \ message("err = %d, size = %d, expected size = %d\n",err,size2,SIZE); \ if (err != 0) { \ if (err != -1) { \ @@ -161,7 +190,7 @@ return; \ } \ \ - if (strcmp(p,VAL) != 0) { \ + if (strcmp(p,val) != 0) { \ fail("value is not correct"); \ return; \ } \ @@ -177,6 +206,51 @@ } \ } \ +#define EI_DECODE_STRING(FUNC,SIZE,VAL) \ + EI_DECODE_STRING_4(FUNC,SIZE,VAL, \ + P99({ERLANG_ANY,ERLANG_ANY,ERLANG_ANY})) + +#define EI_DECODE_STRING_FAIL(FUNC,ATOM) \ + { \ + char p[1024]; \ + char *buf; \ + int size1 = 0; \ + int size2 = 0; \ + int err; \ + struct my_atom atom = ATOM;\ + message("ei_" #FUNC " should fail\n"); \ + p[0] = 0; \ + message("p[0] is %d\n",p[0]); \ + buf = read_packet(NULL); \ +\ + err = ei_ ## FUNC(buf+1, &size1, NULL, &atom); \ + if (err != -1) { \ + fail("should return -1 if NULL pointer"); \ + return; \ + } \ +\ + err = ei_ ## FUNC(buf+1, &size2, p, &atom); \ + if (err != -1) { \ + fail("should return -1"); \ + return; \ + } \ + if (p[0] != 0) { \ + message("p[0] argument was modified to %u\n",(unsigned char)p[0]); \ + } \ +\ + if (size1 != 0) { \ + fail("size of encoded data should be 0 if NULL"); \ + return; \ + } \ +\ + if (size2 != 0) { \ + fail("size of encoded data should be 0"); \ + return; \ + } \ + } \ + +//#define EI_DECODE_UTF8_STRING(FUNC,SIZE,VAL) + #define EI_DECODE_BIN(FUNC,SIZE,VAL,LEN) \ { \ char p[1024]; \ @@ -536,13 +610,13 @@ TESTCASE(test_ei_decode_misc) EI_DECODE_2(decode_boolean, 8, int, 0); EI_DECODE_2(decode_boolean, 7, int, 1); - EI_DECODE_STRING(decode_atom, 6, "foo"); - EI_DECODE_STRING(decode_atom, 3, ""); - EI_DECODE_STRING(decode_atom, 9, "������"); + EI_DECODE_STRING(decode_my_atom, 6, "foo"); + EI_DECODE_STRING(decode_my_atom, 3, ""); + EI_DECODE_STRING(decode_my_atom, 9, "������"); - EI_DECODE_STRING(decode_string, 6, "foo"); - EI_DECODE_STRING(decode_string, 1, ""); - EI_DECODE_STRING(decode_string, 9, "������"); + EI_DECODE_STRING(decode_my_string, 6, "foo"); + EI_DECODE_STRING(decode_my_string, 1, ""); + EI_DECODE_STRING(decode_my_string, 9, "������"); EI_DECODE_BIN(decode_binary, 8, "foo", 3); EI_DECODE_BIN(decode_binary, 5, "", 0); @@ -559,3 +633,63 @@ TESTCASE(test_ei_decode_misc) /* ******************************************************************** */ +TESTCASE(test_ei_decode_utf8_atom) +{ + + EI_DECODE_STRING_4(decode_my_atom_as, 4, P99({229,0}), /* LATIN1 "�" */ + P99({ERLANG_ANY,ERLANG_LATIN1,ERLANG_LATIN1})); + EI_DECODE_STRING_4(decode_my_atom_as, 4, P99({195,164,0}), /* UTF8 "�" */ + P99({ERLANG_UTF8,ERLANG_LATIN1,ERLANG_UTF8})); + EI_DECODE_STRING_4(decode_my_atom_as, 4, P99({246,0}), /* LATIN1 "�" */ + P99({ERLANG_LATIN1,ERLANG_LATIN1,ERLANG_LATIN1})); + EI_DECODE_STRING_FAIL(decode_my_atom_as, + P99({ERLANG_ASCII,ERLANG_ANY,ERLANG_ANY})); + + EI_DECODE_STRING_4(decode_my_atom_as, 4, P99({219,158,0}), + P99({ERLANG_ANY,ERLANG_UTF8,ERLANG_UTF8})); + EI_DECODE_STRING_4(decode_my_atom_as, 6, P99({219,158,219,158,0}), + P99({ERLANG_UTF8,ERLANG_UTF8,ERLANG_UTF8})); + EI_DECODE_STRING_FAIL(decode_my_atom_as, + P99({ERLANG_LATIN1,ERLANG_ANY,ERLANG_ANY})); + EI_DECODE_STRING_FAIL(decode_my_atom_as, + P99({ERLANG_ASCII,ERLANG_ANY,ERLANG_ANY})); + + EI_DECODE_STRING_4(decode_my_atom_as, 4, "a", + P99({ERLANG_ANY,ERLANG_LATIN1,ERLANG_ASCII})); + EI_DECODE_STRING_4(decode_my_atom_as, 4, "b", + P99({ERLANG_UTF8,ERLANG_LATIN1,ERLANG_ASCII})); + EI_DECODE_STRING_4(decode_my_atom_as, 4, "c", + P99({ERLANG_LATIN1,ERLANG_LATIN1,ERLANG_ASCII})); + EI_DECODE_STRING_4(decode_my_atom_as, 4, "d", + P99({ERLANG_ASCII,ERLANG_LATIN1,ERLANG_ASCII})); + + report(1); +} + +/* ******************************************************************** */ + +int ei_decode_my_atom_as(const char *buf, int *index, char *to, + struct my_atom *atom) { + erlang_char_encoding was,result; + int res = ei_decode_atom_as(buf,index,to,1024,atom->from,&was,&result); + if (res != 0) + return res; + if (!(was & atom->was_check)) { + message("Original encoding was %d not %d\n",was,atom->was_check); + return -1; + } else if (!(result & atom->result_check)) { + message("Result encoding was %d not %d\n",result,atom->result_check); + return -1; + } + return res; +} + +int ei_decode_my_atom(const char *buf, int *index, char *to, + struct my_atom *atom) { + return ei_decode_atom(buf, index, to); +} + +int ei_decode_my_string(const char *buf, int *index, char *to, + struct my_atom *atom) { + return ei_decode_string(buf, index, to); +} diff --git a/lib/erl_interface/test/ei_decode_encode_SUITE.erl b/lib/erl_interface/test/ei_decode_encode_SUITE.erl index 85cb62239b..2271278291 100644 --- a/lib/erl_interface/test/ei_decode_encode_SUITE.erl +++ b/lib/erl_interface/test/ei_decode_encode_SUITE.erl @@ -1,7 +1,7 @@ %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 2004-2011. All Rights Reserved. +%% Copyright Ericsson AB 2004-2013. All Rights Reserved. %% %% The contents of this file are subject to the Erlang Public License, %% Version 1.1, (the "License"); you may not use this file except in @@ -118,6 +118,13 @@ test_ei_decode_encode(Config) when is_list(Config) -> ?line send_rec(P, OXPort), ?line send_rec(P, OXRef), + %% Unicode atoms + [begin send_rec(P, Atom), + send_rec(P, mk_pid({Atom,1}, 23434, 3434)), + send_rec(P, mk_port({Atom,1}, 2343434)), + send_rec(P, mk_ref({Atom,1}, [262143, 8723648, 24097245])), + void + end || Atom <- unicode_atom_data()], ?line runner:recv_eot(P), ok. @@ -127,7 +134,7 @@ test_ei_decode_encode(Config) when is_list(Config) -> % We read two packets for each test, the ei_decode_encode and ei_x_decode_encode version.... send_rec(P, Term) when is_port(P) -> - ?t:format("Testing: ~p~n", [Term]), + %%?t:format("Testing: ~p~n", [Term]), P ! {self(), {command, term_to_binary(Term)}}, {_B,Term} = get_buf_and_term(P). @@ -146,7 +153,7 @@ get_buf_and_term(P) -> _ -> B1 = list_to_binary([131,B]), % No magic, add T = binary_to_term(B1), - io:format("~w\n~w\n(got no magic)\n",[B,T]), + %io:format("~w\n~w\n(got no magic)\n",[B,T]), {B,T} end. @@ -160,7 +167,7 @@ get_binary(P) -> case runner:get_term(P) of {bytes,L} -> B = list_to_binary(L), - io:format("~w\n",[L]), + %%io:format("~w\n",[L]), % For strange reasons <<131>> show up as <>.... % io:format("~w\n",[B]), B; @@ -226,38 +233,36 @@ uint8(Uint) -> mk_pid({NodeName, Creation}, Number, Serial) when is_atom(NodeName) -> - mk_pid({atom_to_list(NodeName), Creation}, Number, Serial); -mk_pid({NodeName, Creation}, Number, Serial) -> + <<?VERSION_MAGIC, NodeNameExt/binary>> = term_to_binary(NodeName), + mk_pid({NodeNameExt, Creation}, Number, Serial); +mk_pid({NodeNameExt, Creation}, Number, Serial) -> case catch binary_to_term(list_to_binary([?VERSION_MAGIC, ?PID_EXT, - ?ATOM_EXT, - uint16_be(length(NodeName)), - NodeName, + NodeNameExt, uint32_be(Number), uint32_be(Serial), uint8(Creation)])) of Pid when is_pid(Pid) -> Pid; {'EXIT', {badarg, _}} -> - exit({badarg, mk_pid, [{NodeName, Creation}, Number, Serial]}); + exit({badarg, mk_pid, [{NodeNameExt, Creation}, Number, Serial]}); Other -> exit({unexpected_binary_to_term_result, Other}) end. mk_port({NodeName, Creation}, Number) when is_atom(NodeName) -> - mk_port({atom_to_list(NodeName), Creation}, Number); -mk_port({NodeName, Creation}, Number) -> + <<?VERSION_MAGIC, NodeNameExt/binary>> = term_to_binary(NodeName), + mk_port({NodeNameExt, Creation}, Number); +mk_port({NodeNameExt, Creation}, Number) -> case catch binary_to_term(list_to_binary([?VERSION_MAGIC, ?PORT_EXT, - ?ATOM_EXT, - uint16_be(length(NodeName)), - NodeName, + NodeNameExt, uint32_be(Number), uint8(Creation)])) of Port when is_port(Port) -> Port; {'EXIT', {badarg, _}} -> - exit({badarg, mk_port, [{NodeName, Creation}, Number]}); + exit({badarg, mk_port, [{NodeNameExt, Creation}, Number]}); Other -> exit({unexpected_binary_to_term_result, Other}) end. @@ -265,33 +270,30 @@ mk_port({NodeName, Creation}, Number) -> mk_ref({NodeName, Creation}, Numbers) when is_atom(NodeName), is_integer(Creation), is_list(Numbers) -> - mk_ref({atom_to_list(NodeName), Creation}, Numbers); -mk_ref({NodeName, Creation}, [Number]) when is_list(NodeName), - is_integer(Creation), - is_integer(Number) -> + <<?VERSION_MAGIC, NodeNameExt/binary>> = term_to_binary(NodeName), + mk_ref({NodeNameExt, Creation}, Numbers); +mk_ref({NodeNameExt, Creation}, [Number]) when is_binary(NodeNameExt), + is_integer(Creation), + is_integer(Number) -> case catch binary_to_term(list_to_binary([?VERSION_MAGIC, ?REFERENCE_EXT, - ?ATOM_EXT, - uint16_be(length(NodeName)), - NodeName, + NodeNameExt, uint32_be(Number), uint8(Creation)])) of Ref when is_reference(Ref) -> Ref; {'EXIT', {badarg, _}} -> - exit({badarg, mk_ref, [{NodeName, Creation}, [Number]]}); + exit({badarg, mk_ref, [{NodeNameExt, Creation}, [Number]]}); Other -> exit({unexpected_binary_to_term_result, Other}) end; -mk_ref({NodeName, Creation}, Numbers) when is_list(NodeName), - is_integer(Creation), - is_list(Numbers) -> +mk_ref({NodeNameExt, Creation}, Numbers) when is_binary(NodeNameExt), + is_integer(Creation), + is_list(Numbers) -> case catch binary_to_term(list_to_binary([?VERSION_MAGIC, ?NEW_REFERENCE_EXT, uint16_be(length(Numbers)), - ?ATOM_EXT, - uint16_be(length(NodeName)), - NodeName, + NodeNameExt, uint8(Creation), lists:map(fun (N) -> uint32_be(N) @@ -300,8 +302,67 @@ mk_ref({NodeName, Creation}, Numbers) when is_list(NodeName), Ref when is_reference(Ref) -> Ref; {'EXIT', {badarg, _}} -> - exit({badarg, mk_ref, [{NodeName, Creation}, Numbers]}); + exit({badarg, mk_ref, [{NodeNameExt, Creation}, Numbers]}); Other -> exit({unexpected_binary_to_term_result, Other}) end. + + +unicode_atom_data() -> + [uc_atup(lists:seq(16#1f600, 16#1f600+254)), + uc_atup(lists:seq(16#1f600, 16#1f600+63)), + uc_atup(lists:seq(1, 255)), + uc_atup(lists:seq(100, 163)), + uc_atup(lists:seq(200, 354)), + uc_atup(lists:seq(200, 263)), + uc_atup(lists:seq(2000, 2254)), + uc_atup(lists:seq(2000, 2063)), + uc_atup(lists:seq(65500, 65754)), + uc_atup(lists:seq(65500, 65563)) + | lists:map(fun (N) -> + Pow2 = (1 bsl N), + uc_atup(lists:seq(Pow2 - 127, Pow2 + 127)) + end, + lists:seq(7, 20)) + ]. + +uc_atup(ATxt) -> + string_to_atom(ATxt). + +string_to_atom(String) -> + Utf8List = string_to_utf8_list(String), + Len = length(Utf8List), + TagLen = case Len < 256 of + true -> [119, Len]; + false -> [118, Len bsr 8, Len band 16#ff] + end, + binary_to_term(list_to_binary([131, TagLen, Utf8List])). + +string_to_utf8_list([]) -> + []; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 0 =< CP, + CP =< 16#7F -> + [CP | string_to_utf8_list(CPs)]; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 16#80 =< CP, + CP =< 16#7FF -> + [16#C0 bor (CP bsr 6), + 16#80 bor (16#3F band CP) + | string_to_utf8_list(CPs)]; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 16#800 =< CP, + CP =< 16#FFFF -> + [16#E0 bor (CP bsr 12), + 16#80 bor (16#3F band (CP bsr 6)), + 16#80 bor (16#3F band CP) + | string_to_utf8_list(CPs)]; +string_to_utf8_list([CP|CPs]) when is_integer(CP), + 16#10000 =< CP, + CP =< 16#10FFFF -> + [16#F0 bor (CP bsr 18), + 16#80 bor (16#3F band (CP bsr 12)), + 16#80 bor (16#3F band (CP bsr 6)), + 16#80 bor (16#3F band CP) + | string_to_utf8_list(CPs)]. diff --git a/lib/erl_interface/test/ei_decode_encode_SUITE_data/ei_decode_encode_test.c b/lib/erl_interface/test/ei_decode_encode_SUITE_data/ei_decode_encode_test.c index 406f02ecfb..317e5edecd 100644 --- a/lib/erl_interface/test/ei_decode_encode_SUITE_data/ei_decode_encode_test.c +++ b/lib/erl_interface/test/ei_decode_encode_SUITE_data/ei_decode_encode_test.c @@ -1,7 +1,7 @@ /* * %CopyrightBegin% * - * Copyright Ericsson AB 2004-2009. All Rights Reserved. + * Copyright Ericsson AB 2004-2013. All Rights Reserved. * * The contents of this file are subject to the Erlang Public License, * Version 1.1, (the "License"); you may not use this file except in @@ -29,171 +29,258 @@ * Author: [email protected] */ -#define EI_DECODE_ENCODE(FUNC,TYPE) \ - { \ - char *buf; \ - char buf2[1024]; \ - TYPE p; \ - int size1 = 0; \ - int size2 = 0; \ - int size3 = 0; \ - int err; \ - ei_x_buff arg; \ -\ - message("ei_decode_" #FUNC ", arg is type " #TYPE); \ - buf = read_packet(NULL); \ - err = ei_decode_ ## FUNC(buf+1, &size1, &p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("decode returned non zero but not -1"); \ - } else { \ - fail("decode returned non zero"); \ - } \ - return; \ - } \ - if (size1 < 1) { \ - fail("size is < 1"); \ - return; \ - } \ -\ - message("ei_encode_" #FUNC " buf is NULL, arg is type " #TYPE); \ - err = ei_encode_ ## FUNC(NULL, &size2, &p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("size calculation returned non zero but not -1"); \ - return; \ - } else { \ - fail("size calculation returned non zero"); \ - return; \ - } \ - } \ - if (size1 != size2) { \ - message("size1 = %d, size2 = %d\n",size1,size2); \ - fail("decode and encode size differs when buf is NULL"); \ - return; \ - } \ - message("ei_encode_" #FUNC ", arg is type " #TYPE); \ - err = ei_encode_ ## FUNC(buf2, &size3, &p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("returned non zero but not -1"); \ - } else { \ - fail("returned non zero"); \ - } \ - return; \ - } \ - if (size1 != size3) { \ - message("size1 = %d, size2 = %d\n",size1,size3); \ - fail("decode and encode size differs"); \ - return; \ - } \ - send_buffer(buf2, size1); \ -\ - message("ei_x_encode_" #FUNC ", arg is type " #TYPE); \ - ei_x_new(&arg); \ - err = ei_x_encode_ ## FUNC(&arg, &p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("returned non zero but not -1"); \ - } else { \ - fail("returned non zero"); \ - } \ - ei_x_free(&arg); \ - return; \ - } \ - if (arg.index < 1) { \ - fail("size is < 1"); \ - ei_x_free(&arg); \ - return; \ - } \ - send_buffer(arg.buff, arg.index); \ - ei_x_free(&arg); \ - } - -#define EI_DECODE_ENCODE_BIG(FUNC,TYPE) \ - { \ - char *buf; \ - char buf2[2048]; \ - TYPE *p; \ - int size1 = 0; \ - int size2 = 0; \ - int size3 = 0; \ - int err, index = 0, len, type; \ - ei_x_buff arg; \ -\ - message("ei_decode_" #FUNC ", arg is type " #TYPE); \ - buf = read_packet(NULL); \ - ei_get_type(buf+1, &index, &type, &len); \ - p = ei_alloc_big(len); \ - err = ei_decode_ ## FUNC(buf+1, &size1, p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("decode returned non zero but not -1"); \ - } else { \ - fail("decode returned non zero"); \ - } \ - return; \ - } \ - if (size1 < 1) { \ - fail("size is < 1"); \ - return; \ - } \ -\ - message("ei_encode_" #FUNC " buf is NULL, arg is type " #TYPE); \ - err = ei_encode_ ## FUNC(NULL, &size2, p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("size calculation returned non zero but not -1"); \ - return; \ - } else { \ - fail("size calculation returned non zero"); \ - return; \ - } \ - } \ - if (size1 != size2) { \ - message("size1 = %d, size2 = %d\n",size1,size2); \ - fail("decode and encode size differs when buf is NULL"); \ - return; \ - } \ - message("ei_encode_" #FUNC ", arg is type " #TYPE); \ - err = ei_encode_ ## FUNC(buf2, &size3, p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("returned non zero but not -1"); \ - } else { \ - fail("returned non zero"); \ - } \ - return; \ - } \ - if (size1 != size3) { \ - message("size1 = %d, size2 = %d\n",size1,size3); \ - fail("decode and encode size differs"); \ - return; \ - } \ - send_buffer(buf2, size1); \ -\ - message("ei_x_encode_" #FUNC ", arg is type " #TYPE); \ - ei_x_new(&arg); \ - err = ei_x_encode_ ## FUNC(&arg, p); \ - if (err != 0) { \ - if (err != -1) { \ - fail("returned non zero but not -1"); \ - } else { \ - fail("returned non zero"); \ - } \ - ei_x_free(&arg); \ - return; \ - } \ - if (arg.index < 1) { \ - fail("size is < 1"); \ - ei_x_free(&arg); \ - return; \ - } \ - send_buffer(arg.buff, arg.index); \ - ei_x_free(&arg); \ - ei_free_big(p); \ - } +/*#define MESSAGE(FMT,A1,A2) message(FMT,A1,A2)*/ +#define MESSAGE(FMT,A1,A2) +typedef int decodeFT(const char *buf, int *index, void*); +typedef int encodeFT(char *buf, int *index, void*); +typedef int x_encodeFT(ei_x_buff*, void*); + +struct Type { + char* name; + char* type; + decodeFT* ei_decode_fp; + encodeFT* ei_encode_fp; + x_encodeFT* ei_x_encode_fp; +}; + +typedef struct +{ + char name[MAXATOMLEN_UTF8]; + erlang_char_encoding enc; +}my_atom; + +int ei_decode_my_atom(const char *buf, int *index, my_atom* a) +{ + return ei_decode_atom_as(buf, index, (a ? a->name : NULL), sizeof(a->name), + ERLANG_UTF8, (a ? &a->enc : NULL), NULL); +} +int ei_encode_my_atom(char *buf, int *index, my_atom* a) +{ + return ei_encode_atom_as(buf, index, a->name, ERLANG_UTF8, a->enc); +} +int ei_x_encode_my_atom(ei_x_buff* x, my_atom* a) +{ + return ei_x_encode_atom_as(x, a->name, ERLANG_UTF8, a->enc); +} + +#define BUFSZ 2000 + +void decode_encode(struct Type* t, void* obj) +{ + char *buf; + char buf2[BUFSZ]; + int size1 = 0; + int size2 = 0; + int size3 = 0; + int err; + ei_x_buff arg; + + MESSAGE("ei_decode_%s, arg is type %s", t->name, t->type); + buf = read_packet(NULL); + err = t->ei_decode_fp(buf+1, &size1, NULL); + if (err != 0) { + if (err != -1) { + fail("decode returned non zero but not -1"); + } else { + fail("decode returned non zero"); + } + return; + } + if (size1 < 1) { + fail("size is < 1"); + return; + } + + if (size1 > BUFSZ) { + fail("size is > BUFSZ"); + return; + } + + err = t->ei_decode_fp(buf+1, &size2, obj); + if (err != 0) { + if (err != -1) { + fail("decode returned non zero but not -1"); + } else { + fail("decode returned non zero"); + } + return; + } + if (size1 != size2) { + MESSAGE("size1 = %d, size2 = %d\n",size1,size2); + fail("decode sizes differs"); + return; + } + + size2 = 0; + err = ei_skip_term(buf+1, &size2); + if (err != 0) { + fail("ei_skip_term returned non zero"); + return; + } + if (size1 != size2) { + MESSAGE("size1 = %d, size2 = %d\n",size1,size2); + fail("skip size differs"); + return; + } + + MESSAGE("ei_encode_%s buf is NULL, arg is type %s", t->name, t->type); + size2 = 0; + err = t->ei_encode_fp(NULL, &size2, obj); + if (err != 0) { + if (err != -1) { + fail("size calculation returned non zero but not -1"); + return; + } else { + fail("size calculation returned non zero"); + return; + } + } + if (size1 != size2) { + MESSAGE("size1 = %d, size2 = %d\n",size1,size2); + fail("decode and encode size differs when buf is NULL"); + return; + } + MESSAGE("ei_encode_%s, arg is type %s", t->name, t->type); + err = t->ei_encode_fp(buf2, &size3, obj); + if (err != 0) { + if (err != -1) { + fail("returned non zero but not -1"); + } else { + fail("returned non zero"); + } + return; + } + if (size1 != size3) { + MESSAGE("size1 = %d, size2 = %d\n",size1,size3); + fail("decode and encode size differs"); + return; + } + send_buffer(buf2, size1); + + MESSAGE("ei_x_encode_%s, arg is type %s", t->name, t->type); + ei_x_new(&arg); + err = t->ei_x_encode_fp(&arg, obj); + if (err != 0) { + if (err != -1) { + fail("returned non zero but not -1"); + } else { + fail("returned non zero"); + } + ei_x_free(&arg); + return; + } + if (arg.index < 1) { + fail("size is < 1"); + ei_x_free(&arg); + return; + } + send_buffer(arg.buff, arg.index); + ei_x_free(&arg); +} + + +#define EI_DECODE_ENCODE(TYPE, ERLANG_TYPE) { \ + struct Type type_struct = {#TYPE, #ERLANG_TYPE, \ + (decodeFT*)ei_decode_##TYPE, \ + (encodeFT*)ei_encode_##TYPE, \ + (x_encodeFT*)ei_x_encode_##TYPE }; \ + ERLANG_TYPE type_obj; \ + decode_encode(&type_struct, &type_obj); \ + } + + +void decode_encode_big(struct Type* t) +{ + char *buf; + char buf2[2048]; + void *p; /* (TYPE*) */ + int size1 = 0; + int size2 = 0; + int size3 = 0; + int err, index = 0, len, type; + ei_x_buff arg; + + MESSAGE("ei_decode_%s, arg is type %s", t->name, t->type); + buf = read_packet(NULL); + ei_get_type(buf+1, &index, &type, &len); + p = ei_alloc_big(len); + err = t->ei_decode_fp(buf+1, &size1, p); + if (err != 0) { + if (err != -1) { + fail("decode returned non zero but not -1"); + } else { + fail("decode returned non zero"); + } + return; + } + if (size1 < 1) { + fail("size is < 1"); + return; + } + + MESSAGE("ei_encode_%s buf is NULL, arg is type %s", t->name, t->type); + err = t->ei_encode_fp(NULL, &size2, p); + if (err != 0) { + if (err != -1) { + fail("size calculation returned non zero but not -1"); + return; + } else { + fail("size calculation returned non zero"); + return; + } + } + if (size1 != size2) { + MESSAGE("size1 = %d, size2 = %d\n",size1,size2); + fail("decode and encode size differs when buf is NULL"); + return; + } + MESSAGE("ei_encode_%s, arg is type %s", t->name, t->type); + err = t->ei_encode_fp(buf2, &size3, p); + if (err != 0) { + if (err != -1) { + fail("returned non zero but not -1"); + } else { + fail("returned non zero"); + } + return; + } + if (size1 != size3) { + MESSAGE("size1 = %d, size2 = %d\n",size1,size3); + fail("decode and encode size differs"); + return; + } + send_buffer(buf2, size1); + + MESSAGE("ei_x_encode_%s, arg is type %s", t->name, t->type); + ei_x_new(&arg); + err = t->ei_x_encode_fp(&arg, p); + if (err != 0) { + if (err != -1) { + fail("returned non zero but not -1"); + } else { + fail("returned non zero"); + } + ei_x_free(&arg); + return; + } + if (arg.index < 1) { + fail("size is < 1"); + ei_x_free(&arg); + return; + } + send_buffer(arg.buff, arg.index); + ei_x_free(&arg); + ei_free_big(p); +} + +#define EI_DECODE_ENCODE_BIG(TYPE, ERLANG_TYPE) { \ + struct Type type_struct = {#TYPE, #ERLANG_TYPE, \ + (decodeFT*)ei_decode_##TYPE, \ + (encodeFT*)ei_encode_##TYPE, \ + (x_encodeFT*)ei_x_encode_##TYPE }; \ + decode_encode_big(&type_struct); \ + } @@ -201,6 +288,8 @@ TESTCASE(test_ei_decode_encode) { + int i; + EI_DECODE_ENCODE(fun , erlang_fun); EI_DECODE_ENCODE(pid , erlang_pid); EI_DECODE_ENCODE(port , erlang_port); @@ -223,6 +312,14 @@ TESTCASE(test_ei_decode_encode) EI_DECODE_ENCODE(port , erlang_port); EI_DECODE_ENCODE(ref , erlang_ref); + /* Unicode atoms */ + for (i=0; i<24; i++) { + EI_DECODE_ENCODE(my_atom, my_atom); + EI_DECODE_ENCODE(pid, erlang_pid); + EI_DECODE_ENCODE(port, erlang_port); + EI_DECODE_ENCODE(ref, erlang_ref); + } + report(1); } diff --git a/lib/erl_interface/test/ei_encode_SUITE.erl b/lib/erl_interface/test/ei_encode_SUITE.erl index cefd33e5f6..e0ac4c0637 100644 --- a/lib/erl_interface/test/ei_encode_SUITE.erl +++ b/lib/erl_interface/test/ei_encode_SUITE.erl @@ -1,7 +1,8 @@ +%% -*- coding: utf-8 -*- %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 2004-2011. All Rights Reserved. +%% Copyright Ericsson AB 2004-2012. All Rights Reserved. %% %% The contents of this file are subject to the Erlang Public License, %% Version 1.1, (the "License"); you may not use this file except in @@ -33,7 +34,9 @@ test_ei_encode_ulonglong/1, test_ei_encode_char/1, test_ei_encode_misc/1, - test_ei_encode_fails/1 + test_ei_encode_fails/1, + test_ei_encode_utf8_atom/1, + test_ei_encode_utf8_atom_len/1 ]). suite() -> [{ct_hooks,[ts_install_cth]}]. @@ -42,7 +45,8 @@ all() -> [test_ei_encode_long, test_ei_encode_ulong, test_ei_encode_longlong, test_ei_encode_ulonglong, test_ei_encode_char, test_ei_encode_misc, - test_ei_encode_fails]. + test_ei_encode_fails, test_ei_encode_utf8_atom, + test_ei_encode_utf8_atom_len]. groups() -> []. @@ -213,19 +217,19 @@ test_ei_encode_misc(Config) when is_list(Config) -> ?line {<<100,0,3,"foo">>,foo} = get_buf_and_term(P), ?line {<<100,0,0,"">>,''} = get_buf_and_term(P), ?line {<<100,0,0,"">>,''} = get_buf_and_term(P), - ?line {<<100,0,6,"������">>,'������'} = get_buf_and_term(P), - ?line {<<100,0,6,"������">>,'������'} = get_buf_and_term(P), + ?line {<<100,0,6,"ÅÄÖåäö">>,'ÅÄÖåäö'} = get_buf_and_term(P), + ?line {<<100,0,6,"ÅÄÖåäö">>,'ÅÄÖåäö'} = get_buf_and_term(P), ?line {<<107,0,3,"foo">>,"foo"} = get_buf_and_term(P), ?line {<<107,0,3,"foo">>,"foo"} = get_buf_and_term(P), ?line {<<106>>,""} = get_buf_and_term(P), ?line {<<106>>,""} = get_buf_and_term(P), - ?line {<<107,0,6,"������">>,"������"} = get_buf_and_term(P), - ?line {<<107,0,6,"������">>,"������"} = get_buf_and_term(P), + ?line {<<107,0,6,"ÅÄÖåäö">>,"ÅÄÖåäö"} = get_buf_and_term(P), + ?line {<<107,0,6,"ÅÄÖåäö">>,"ÅÄÖåäö"} = get_buf_and_term(P), ?line {<<109,0,0,0,3,"foo">>,<<"foo">>} = get_buf_and_term(P), ?line {<<109,0,0,0,0,"">>,<<>>} = get_buf_and_term(P), - ?line {<<109,0,0,0,6,"������">>,<<"������">>} = get_buf_and_term(P), + ?line {<<109,0,0,0,6,"ÅÄÖåäö">>,<<"ÅÄÖåäö">>} = get_buf_and_term(P), ?line {<<104,0>>,{}} = get_buf_and_term(P), % Tuple header for {} ?line {<<106>>,[]} = get_buf_and_term(P), % Empty list [] @@ -256,6 +260,38 @@ test_ei_encode_fails(Config) when is_list(Config) -> %% ######################################################################## %% +test_ei_encode_utf8_atom(Config) -> + ?line P = runner:start(?test_ei_encode_utf8_atom), + + ?line {<<119,2,195,133>>,'Å'} = get_buf_and_term(P), + ?line {<<100,0,1,197>>,'Å'} = get_buf_and_term(P), + ?line {<<100,0,1,197>>,'Å'} = get_buf_and_term(P), + ?line {<<119,2,195,133>>,'Å'} = get_buf_and_term(P), + + ?line {<<119,1,$A>>,'A'} = get_buf_and_term(P), + ?line {<<100,0,1,$A>>,'A'} = get_buf_and_term(P), + + ?line runner:recv_eot(P), + ok. + +%% ######################################################################## %% +test_ei_encode_utf8_atom_len(Config) -> + ?line P = runner:start(?test_ei_encode_utf8_atom_len), + + ?line {<<119,2,195,133>>,'Å'} = get_buf_and_term(P), + ?line {<<100,0,2,197,196>>,'ÅÄ'} = get_buf_and_term(P), + ?line {<<100,0,1,197>>,'Å'} = get_buf_and_term(P), + ?line {<<119,4,195,133,195,132>>,'ÅÄ'} = get_buf_and_term(P), + + ?line {<<119,1,$A>>,'A'} = get_buf_and_term(P), + ?line {<<100,0,2,$A,$B>>,'AB'} = get_buf_and_term(P), + ?line {<<100,0,255,_:(255*8)>>,_} = get_buf_and_term(P), + + ?line runner:recv_eot(P), + ok. + +%% ######################################################################## %% + % We read two packets for each test, the ei_encode and ei_x_encode version.... get_buf_and_term(P) -> diff --git a/lib/erl_interface/test/ei_encode_SUITE_data/ei_encode_test.c b/lib/erl_interface/test/ei_encode_SUITE_data/ei_encode_test.c index c373658152..e904375e9e 100644 --- a/lib/erl_interface/test/ei_encode_SUITE_data/ei_encode_test.c +++ b/lib/erl_interface/test/ei_encode_SUITE_data/ei_encode_test.c @@ -208,6 +208,196 @@ ei_x_free(&arg); \ } +#define EI_ENCODE_3(FUNC,ARG1,ARG2,ARG3) \ + { \ + char buf[1024]; \ + int size1 = 0; \ + int size2 = 0; \ + int err; \ + message("ei_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " encoded as "); \ + err = ei_ ## FUNC(NULL, &size1, ARG1, ARG2, ARG3); \ + if (err != 0) { \ + if (err != -1) { \ + fail("size calculation returned non zero but not -1"); \ + return; \ + } else { \ + fail("size calculation returned non zero"); \ + return; \ + } \ + } \ + err = ei_ ## FUNC(buf, &size2, ARG1, ARG2, ARG3); \ + if (err != 0) { \ + if (err != -1) { \ + fail("returned non zero but not -1"); \ + } else { \ + fail("returned non zero"); \ + } \ + return; \ + } \ + if (size1 != size2) { \ + fail("size differs when arg is NULL or buf"); \ + return; \ + } \ + if (size1 < 1) { \ + fail("size is < 1"); \ + return; \ + } \ + send_buffer(buf, size1); \ + } \ + { \ + ei_x_buff arg; \ + int err; \ + message("ei_x_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " encoded as "); \ + ei_x_new(&arg); \ + err = ei_x_ ## FUNC(&arg, ARG1, ARG2, ARG3); \ + if (err != 0) { \ + if (err != -1) { \ + fail("returned non zero but not -1"); \ + } else { \ + fail("returned non zero"); \ + } \ + ei_x_free(&arg); \ + return; \ + } \ + if (arg.index < 1) { \ + fail("size is < 1"); \ + ei_x_free(&arg); \ + return; \ + } \ + send_buffer(arg.buff, arg.index); \ + ei_x_free(&arg); \ + } + +#define EI_ENCODE_3_FAIL(FUNC,ARG1,ARG2,ARG3) \ + { \ + char buf[1024]; \ + int size1 = 0; \ + int size2 = 0; \ + int err; \ + message("ei_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " encoded as "); \ + err = ei_ ## FUNC(NULL, &size1, ARG1, ARG2, ARG3); \ + if (err != -1) { \ + fail("size calculation returned non -1"); \ + return; \ + } \ + err = ei_ ## FUNC(buf, &size2, ARG1, ARG2, ARG3); \ + if (err != -1) { \ + fail("returned non -1"); \ + } \ + if (size1 != size2) { \ + fail("size differs when arg is NULL or buf"); \ + return; \ + } \ + } \ + { \ + ei_x_buff arg; \ + int err; \ + message("ei_x_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " encoded as "); \ + ei_x_new(&arg); \ + err = ei_x_ ## FUNC(&arg, ARG1, ARG2, ARG3); \ + if (err != -1) { \ + fail("returned non -1"); \ + ei_x_free(&arg); \ + return; \ + } \ + ei_x_free(&arg); \ + } + +#define EI_ENCODE_4(FUNC,ARG1,ARG2,ARG3,ARG4) \ + { \ + char buf[1024]; \ + int size1 = 0; \ + int size2 = 0; \ + int err; \ + message("ei_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " " #ARG4 " encoded as "); \ + err = ei_ ## FUNC(NULL, &size1, ARG1, ARG2, ARG3, ARG4); \ + if (err != 0) { \ + if (err != -1) { \ + fail("size calculation returned non zero but not -1"); \ + return; \ + } else { \ + fail("size calculation returned non zero"); \ + return; \ + } \ + } \ + err = ei_ ## FUNC(buf, &size2, ARG1, ARG2, ARG3, ARG4); \ + if (err != 0) { \ + if (err != -1) { \ + fail("returned non zero but not -1"); \ + } else { \ + fail("returned non zero"); \ + } \ + return; \ + } \ + if (size1 != size2) { \ + fail("size differs when arg is NULL or buf"); \ + return; \ + } \ + if (size1 < 1) { \ + fail("size is < 1"); \ + return; \ + } \ + send_buffer(buf, size1); \ + } \ + { \ + ei_x_buff arg; \ + int err; \ + message("ei_x_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " " #ARG4 " encoded as "); \ + ei_x_new(&arg); \ + err = ei_x_ ## FUNC(&arg, ARG1, ARG2, ARG3, ARG4); \ + if (err != 0) { \ + if (err != -1) { \ + fail("returned non zero but not -1"); \ + } else { \ + fail("returned non zero"); \ + } \ + ei_x_free(&arg); \ + return; \ + } \ + if (arg.index < 1) { \ + fail("size is < 1"); \ + ei_x_free(&arg); \ + return; \ + } \ + send_buffer(arg.buff, arg.index); \ + ei_x_free(&arg); \ + } + +#define EI_ENCODE_4_FAIL(FUNC,ARG1,ARG2,ARG3,ARG4) \ + { \ + char buf[1024]; \ + int size1 = 0; \ + int size2 = 0; \ + int err; \ + message("ei_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " " #ARG4 " encoded as "); \ + err = ei_ ## FUNC(NULL, &size1, ARG1, ARG2, ARG3, ARG4); \ + if (err != -1) { \ + fail("size calculation returned non -1"); \ + return; \ + } \ + err = ei_ ## FUNC(buf, &size2, ARG1, ARG2, ARG3, ARG4); \ + if (err != -1) { \ + fail("returned non -1"); \ + } \ + if (size1 != size2) { \ + fail("size differs when arg is NULL or buf"); \ + return; \ + } \ + } \ + { \ + ei_x_buff arg; \ + int err; \ + message("ei_x_" #FUNC " " #ARG1 " " #ARG2 " " #ARG3 " " #ARG4 " encoded as "); \ + ei_x_new(&arg); \ + err = ei_x_ ## FUNC(&arg, ARG1, ARG2, ARG3, ARG4); \ + if (err != -1) { \ + fail("returned non -1"); \ + ei_x_free(&arg); \ + return; \ + } \ + ei_x_free(&arg); \ + } + /* ******************************************************************** */ TESTCASE(test_ei_encode_long) @@ -464,3 +654,57 @@ TESTCASE(test_ei_encode_fails) report(1); } + +/* ******************************************************************** */ + +TESTCASE(test_ei_encode_utf8_atom) +{ + + EI_ENCODE_3(encode_atom_as, "�", ERLANG_LATIN1, ERLANG_UTF8); + EI_ENCODE_3(encode_atom_as, "�", ERLANG_LATIN1, ERLANG_LATIN1); + EI_ENCODE_3(encode_atom_as, "\303\205", + ERLANG_UTF8, ERLANG_LATIN1); + EI_ENCODE_3(encode_atom_as, "\303\205", + ERLANG_UTF8, ERLANG_UTF8); + + EI_ENCODE_3(encode_atom_as, "A", ERLANG_ASCII, ERLANG_UTF8); + EI_ENCODE_3(encode_atom_as, "A", ERLANG_ASCII, ERLANG_LATIN1); + + EI_ENCODE_3_FAIL(encode_atom_as, "�", ERLANG_UTF8, ERLANG_UTF8); + EI_ENCODE_3_FAIL(encode_atom_as, "�", ERLANG_UTF8, ERLANG_LATIN1); + EI_ENCODE_3_FAIL(encode_atom_as, "�", ERLANG_ASCII, ERLANG_UTF8); + EI_ENCODE_3_FAIL(encode_atom_as, "�", ERLANG_ASCII, ERLANG_LATIN1); + + EI_ENCODE_3_FAIL(encode_atom_as, "yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy", ERLANG_ASCII, ERLANG_LATIN1); + EI_ENCODE_3_FAIL(encode_atom_as, "yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy", ERLANG_LATIN1, ERLANG_LATIN1); + + EI_ENCODE_3_FAIL(encode_atom_as, "yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy", ERLANG_UTF8, ERLANG_LATIN1); + + report(1); +} + +TESTCASE(test_ei_encode_utf8_atom_len) +{ + + EI_ENCODE_4(encode_atom_len_as, "���", 1, ERLANG_LATIN1, ERLANG_UTF8); + EI_ENCODE_4(encode_atom_len_as, "���", 2, ERLANG_LATIN1, ERLANG_LATIN1); + EI_ENCODE_4(encode_atom_len_as, "\303\205\303\204\303\226", 2, + ERLANG_UTF8, ERLANG_LATIN1); + EI_ENCODE_4(encode_atom_len_as, "\303\205\303\204\303\226", 4, + ERLANG_UTF8, ERLANG_UTF8); + + EI_ENCODE_4(encode_atom_len_as, "ABC", 1, ERLANG_ASCII, ERLANG_UTF8); + EI_ENCODE_4(encode_atom_len_as, "ABC", 2, ERLANG_ASCII, ERLANG_LATIN1); + + EI_ENCODE_4_FAIL(encode_atom_len_as, "���", 1, ERLANG_UTF8, ERLANG_UTF8); + EI_ENCODE_4_FAIL(encode_atom_len_as, "���", 2, ERLANG_UTF8, ERLANG_LATIN1); + EI_ENCODE_4_FAIL(encode_atom_len_as, "���", 3, ERLANG_ASCII, ERLANG_UTF8); + EI_ENCODE_4_FAIL(encode_atom_len_as, "���", 4, ERLANG_ASCII, ERLANG_LATIN1); + + EI_ENCODE_4(encode_atom_len_as, "yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy", 255, ERLANG_ASCII, ERLANG_LATIN1); + + EI_ENCODE_4_FAIL(encode_atom_len_as, "\303\205\303\204\303\226", 1, + ERLANG_UTF8, ERLANG_LATIN1); + + report(1); +} diff --git a/lib/erl_interface/test/ei_print_SUITE.erl b/lib/erl_interface/test/ei_print_SUITE.erl index 2a3ed81f53..6305302e28 100644 --- a/lib/erl_interface/test/ei_print_SUITE.erl +++ b/lib/erl_interface/test/ei_print_SUITE.erl @@ -1,7 +1,7 @@ %% %% %CopyrightBegin% %% -%% Copyright Ericsson AB 2001-2011. All Rights Reserved. +%% Copyright Ericsson AB 2001-2012. All Rights Reserved. %% %% The contents of this file are subject to the Erlang Public License, %% Version 1.1, (the "License"); you may not use this file except in @@ -136,7 +136,7 @@ lists(Config) when is_list(Config) -> %% ?line {term, "[{name, 'Madonna'}, {age, 21}, {data, [{addr, "E-street", 42}]}]"} = %% get_term(P), - %% kanske regexp i st�llet? + %% maybe regexp instead? ?line {term, "[{pi, 3.141500}, {'cos(70)', 0.342020}]"} = get_term(P), ?line {term, "[[pi, 3.141500], ['cos(70)', 0.342020]]"} = get_term(P), diff --git a/lib/erl_interface/vsn.mk b/lib/erl_interface/vsn.mk index 1718f38069..2a3843bda6 100644 --- a/lib/erl_interface/vsn.mk +++ b/lib/erl_interface/vsn.mk @@ -1 +1 @@ -EI_VSN = 3.7.9 +EI_VSN = 3.7.10 |