aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorPatrik Nyblom <[email protected]>2013-07-18 10:18:58 +0200
committerPatrik Nyblom <[email protected]>2013-08-09 12:10:30 +0200
commit6146e7642d4bb9f7c9bb5f8cbca548c1d9667e5c (patch)
tree024c55cef26cdf0ad167b23d61fee9737177da7b
parent9cd8b5d2af163f29cf77ae74057789be977f6414 (diff)
downloadotp-6146e7642d4bb9f7c9bb5f8cbca548c1d9667e5c.tar.gz
otp-6146e7642d4bb9f7c9bb5f8cbca548c1d9667e5c.tar.bz2
otp-6146e7642d4bb9f7c9bb5f8cbca548c1d9667e5c.zip
Add new options to Erlang re interface and mend dupnames
Add notempty_atstart, no_start_optimize, ucp and never_utf options from new PCRE version. Use the new notempty_atstart in global matching. Add inspect/2 function Correctly handle dupnames when capturing a name, as in Perl, get the leftmost matching occurence. Also added all_names, to get all the names in the pattern in alphabetical (name) order. To be able to use this in global matching, an inspect function that can dig out a namelist was added.
-rw-r--r--erts/emulator/beam/atom.names6
-rw-r--r--erts/emulator/beam/bif.tab6
-rw-r--r--erts/emulator/beam/erl_bif_re.c233
-rw-r--r--lib/stdlib/src/re.erl72
-rw-r--r--lib/stdlib/test/re_SUITE.erl159
5 files changed, 446 insertions, 30 deletions
diff --git a/erts/emulator/beam/atom.names b/erts/emulator/beam/atom.names
index eba1d0fa23..cf8f511b85 100644
--- a/erts/emulator/beam/atom.names
+++ b/erts/emulator/beam/atom.names
@@ -71,6 +71,7 @@ atom ac
atom active
atom all
atom all_but_first
+atom all_names
atom alloc_info
atom alloc_sizes
atom allocated
@@ -348,11 +349,13 @@ atom multi_scheduling
atom multiline
atom name
atom named_table
+atom namelist
atom native_addresses
atom Neq='=/='
atom Neqeq='/='
atom net_kernel
atom net_kernel_terminated
+atom never_utf
atom new
atom new_index
atom new_uniq
@@ -378,6 +381,7 @@ atom nosuspend
atom no_float
atom no_integer
atom no_network
+atom no_start_optimize
atom not
atom not_a_list
atom not_loaded
@@ -388,6 +392,7 @@ atom notalive
atom notbol
atom noteol
atom notempty
+atom notempty_atstart
atom notify
atom notsup
atom nouse_stdio
@@ -554,6 +559,7 @@ atom true
atom tuple
atom type
atom ucompile
+atom ucp
atom undef
atom ungreedy
atom unicode
diff --git a/erts/emulator/beam/bif.tab b/erts/emulator/beam/bif.tab
index dc8e9101de..7c8e4b31cf 100644
--- a/erts/emulator/beam/bif.tab
+++ b/erts/emulator/beam/bif.tab
@@ -574,6 +574,12 @@ bif erlang:binary_to_float/1
bif io:printable_range/0
#
+# New in R17A
+#
+
+bif re:inspect/2
+
+#
# Obsolete
#
diff --git a/erts/emulator/beam/erl_bif_re.c b/erts/emulator/beam/erl_bif_re.c
index 12fc834685..c74125ae41 100644
--- a/erts/emulator/beam/erl_bif_re.c
+++ b/erts/emulator/beam/erl_bif_re.c
@@ -288,6 +288,10 @@ parse_options(Eterm listp, /* in */
eopt |= PCRE_NOTEMPTY;
fl |= PARSE_FLAG_UNIQUE_EXEC_OPT;
break;
+ case am_notempty_atstart:
+ eopt |= PCRE_NOTEMPTY_ATSTART;
+ fl |= PARSE_FLAG_UNIQUE_EXEC_OPT;
+ break;
case am_notbol:
eopt |= PCRE_NOTBOL;
fl |= PARSE_FLAG_UNIQUE_EXEC_OPT;
@@ -296,6 +300,10 @@ parse_options(Eterm listp, /* in */
eopt |= PCRE_NOTEOL;
fl |= PARSE_FLAG_UNIQUE_EXEC_OPT;
break;
+ case am_no_start_optimize:
+ copt |= PCRE_NO_START_OPTIMIZE;
+ fl |= PARSE_FLAG_UNIQUE_COMPILE_OPT;
+ break;
case am_caseless:
copt |= PCRE_CASELESS;
fl |= PARSE_FLAG_UNIQUE_COMPILE_OPT;
@@ -332,6 +340,14 @@ parse_options(Eterm listp, /* in */
copt |= PCRE_UNGREEDY;
fl |= PARSE_FLAG_UNIQUE_COMPILE_OPT;
break;
+ case am_ucp:
+ copt |= PCRE_UCP;
+ fl |= PARSE_FLAG_UNIQUE_COMPILE_OPT;
+ break;
+ case am_never_utf:
+ copt |= PCRE_NEVER_UTF;
+ fl |= PARSE_FLAG_UNIQUE_COMPILE_OPT;
+ break;
case am_unicode:
copt |= PCRE_UTF8;
fl |= (PARSE_FLAG_UNIQUE_COMPILE_OPT | PARSE_FLAG_UNICODE);
@@ -359,7 +375,7 @@ parse_options(Eterm listp, /* in */
if (compile_options != NULL) {
*compile_options = copt;
}
- if (exec_options != NULL) {
+ if (exec_options != NULL) {
*exec_options = eopt;
}
if (flags != NULL) {
@@ -585,6 +601,17 @@ static Eterm build_exec_return(Process *p, int rc, RestartContext *restartp, Ete
ri->num_spec * 2 * sizeof(Eterm));
for (i = 0; i < ri->num_spec; ++i) {
x = ri->v[i];
+ if (x < -1) {
+ int n = i-x+1;
+ int j;
+ for (j = i+1; j < ri->num_spec && j < n; ++j) {
+ if (restartp->ovector[(ri->v[j])*2] >= 0) {
+ x = ri->v[j];
+ break;
+ }
+ }
+ i = n-1;
+ }
if (x < rc && x >= 0) {
tmp_vect[n*2] = make_signed_integer(restartp->ovector[x*2],p);
tmp_vect[n*2+1] = make_signed_integer(restartp->ovector[x*2+1]-restartp->ovector[x*2],p);
@@ -666,6 +693,17 @@ static Eterm build_exec_return(Process *p, int rc, RestartContext *restartp, Ete
ri->num_spec * sizeof(Eterm));
for (i = 0; i < ri->num_spec; ++i) {
x = ri->v[i];
+ if (x < -1) {
+ int n = i-x+1;
+ int j;
+ for (j = i+1; j < ri->num_spec && j < n; ++j) {
+ if (restartp->ovector[(ri->v[j])*2] >= 0) {
+ x = ri->v[j];
+ break;
+ }
+ }
+ i = n-1;
+ }
if (x < rc && x >= 0) {
char *cp;
int len;
@@ -730,6 +768,49 @@ static Eterm build_exec_return(Process *p, int rc, RestartContext *restartp, Ete
*/
#define RINFO_SIZ(Num) (sizeof(ReturnInfo) + (sizeof(int) * (Num - 1)))
+#define PICK_INDEX(NameEntry) \
+ ((int) ((((unsigned) ((unsigned char *) (NameEntry))[0]) << 8) + \
+ ((unsigned) ((unsigned char *) (NameEntry))[1])))
+
+
+static void build_one_capture(const pcre *code, ReturnInfo **ri, int *sallocated, int has_dupnames, char *name)
+{
+ ReturnInfo *r = (*ri);
+ if (has_dupnames) {
+ /* Build a sequence of positions, starting with -size if
+ more than one, otherwise just put the index there... */
+ char *first,*last;
+ int esize = erts_pcre_get_stringtable_entries(code,name,&first,&last);
+ if (esize == PCRE_ERROR_NOSUBSTRING) {
+ r->v[r->num_spec - 1] = -1;
+ } else if(last == first) {
+ r->v[r->num_spec - 1] = PICK_INDEX(first);
+ } else {
+ int num = ((last - first) / esize) + 1;
+ int i;
+ ASSERT(num > 1);
+ r->v[r->num_spec - 1] = -num; /* A value less than -1 means
+ multiple indexes for same name */
+ for (i = 0; i < num; ++i) {
+ ++(r->num_spec);
+ if(r->num_spec > (*sallocated)) {
+ (*sallocated) += 10;
+ r = erts_realloc(ERTS_ALC_T_RE_SUBJECT, r,
+ RINFO_SIZ((*sallocated)));
+ }
+ r->v[r->num_spec - 1] = PICK_INDEX(first);
+ first += esize;
+ }
+ }
+ } else {
+ /* Use the faster binary search if no duplicate names are present */
+ if ((r->v[r->num_spec - 1] = erts_pcre_get_stringnumber(code,name)) ==
+ PCRE_ERROR_NOSUBSTRING) {
+ r->v[r->num_spec - 1] = -1;
+ }
+ }
+ *ri = r;
+}
static ReturnInfo *
build_capture(Eterm capture_spec[CAPSPEC_SIZE], const pcre *code)
@@ -778,6 +859,53 @@ build_capture(Eterm capture_spec[CAPSPEC_SIZE], const pcre *code)
}
ri->v[ri->num_spec - 1] = 0;
break;
+ case am_all_names:
+ {
+ int rc,i,top;
+ int entrysize;
+ char *nametable, *last = NULL;
+ int has_dupnames;
+ unsigned long options;
+
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_OPTIONS, &options) != 0)
+ goto error;
+ if ((rc = erts_pcre_fullinfo(code, NULL, PCRE_INFO_NAMECOUNT, &top)) != 0)
+ goto error;
+ if (top <= 0) {
+ ri->num_spec = 0;
+ ri->type = RetNone;
+ break;
+ }
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_NAMEENTRYSIZE, &entrysize) != 0)
+ goto error;
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_NAMETABLE, (unsigned char **) &nametable) != 0)
+ goto error;
+
+ has_dupnames = ((options & PCRE_DUPNAMES) != 0);
+
+ for(i=0;i<top;++i) {
+ if (last == NULL || !has_dupnames || strcmp(last+2,nametable+2)) {
+ if (ri->num_spec < 0)
+ ri->num_spec = 0;
+ ++(ri->num_spec);
+ if(ri->num_spec > sallocated) {
+ sallocated += 10;
+ ri = erts_realloc(ERTS_ALC_T_RE_SUBJECT, ri, RINFO_SIZ(sallocated));
+ }
+ if (has_dupnames) {
+ /* This could be more effective, we actually have
+ the names and could fill in the vector
+ immediately. Now we lookup the name again. */
+ build_one_capture(code,&ri,&sallocated,has_dupnames,nametable+2);
+ } else {
+ ri->v[ri->num_spec - 1] = PICK_INDEX(nametable);
+ }
+ }
+ last = nametable;
+ nametable += entrysize;
+ }
+ break;
+ }
default:
if (is_list(capture_spec[CAPSPEC_VALUES])) {
for(l=capture_spec[CAPSPEC_VALUES];is_list(l);l = CDR(list_val(l))) {
@@ -793,6 +921,11 @@ build_capture(Eterm capture_spec[CAPSPEC_SIZE], const pcre *code)
if (term_to_int(val,&x)) {
ri->v[ri->num_spec - 1] = x;
} else if (is_atom(val) || is_binary(val) || is_list(val)) {
+ int has_dupnames;
+ unsigned long options;
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_OPTIONS, &options) != 0)
+ goto error;
+ has_dupnames = ((options & PCRE_DUPNAMES) != 0);
if (is_atom(val)) {
Atom *ap = atom_tab(atom_val(val));
if ((ap->len + 1) > tmpbsiz) {
@@ -823,10 +956,7 @@ build_capture(Eterm capture_spec[CAPSPEC_SIZE], const pcre *code)
}
tmpb[slen] = '\0';
}
- if ((ri->v[ri->num_spec - 1] = erts_pcre_get_stringnumber(code,tmpb)) ==
- PCRE_ERROR_NOSUBSTRING) {
- ri->v[ri->num_spec - 1] = -1;
- }
+ build_one_capture(code,&ri,&sallocated,has_dupnames,tmpb);
} else {
goto error;
}
@@ -1159,6 +1289,99 @@ static BIF_RETTYPE re_exec_trap(BIF_ALIST_3)
BIF_RET(res);
}
+BIF_RETTYPE
+re_inspect_2(BIF_ALIST_2)
+{
+ Eterm *tp,*tmp_vec,*hp;
+ int rc,i,top,j;
+ int entrysize;
+ char *nametable, *last,*name;
+ int has_dupnames;
+ unsigned long options;
+ int num_names;
+ Eterm res;
+ const pcre *code;
+ byte *temp_alloc = NULL;
+
+ if (is_not_tuple(BIF_ARG_1) || (arityval(*tuple_val(BIF_ARG_1)) != 5)) {
+ goto error;
+ }
+ tp = tuple_val(BIF_ARG_1);
+ if (tp[1] != am_re_pattern || is_not_small(tp[2]) ||
+ is_not_small(tp[3]) || is_not_small(tp[4]) ||
+ is_not_binary(tp[5])) {
+ goto error;
+ }
+ if (BIF_ARG_2 != am_namelist) {
+ goto error;
+ }
+ if ((code = (const pcre *)
+ erts_get_aligned_binary_bytes(tp[5], &temp_alloc)) == NULL) {
+ goto error;
+ }
+
+ /* OK, so let's try to get some info */
+
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_OPTIONS, &options) != 0)
+ goto error;
+ if ((rc = erts_pcre_fullinfo(code, NULL, PCRE_INFO_NAMECOUNT, &top)) != 0)
+ goto error;
+ if (top <= 0) {
+ hp = HAlloc(BIF_P, 3);
+ res = TUPLE2(hp,am_namelist,NIL);
+ erts_free_aligned_binary_bytes(temp_alloc);
+ BIF_RET(res);
+ }
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_NAMEENTRYSIZE, &entrysize) != 0)
+ goto error;
+ if (erts_pcre_fullinfo(code, NULL, PCRE_INFO_NAMETABLE, (unsigned char **) &nametable) != 0)
+ goto error;
+
+ has_dupnames = ((options & PCRE_DUPNAMES) != 0);
+ /* First, count the names */
+ num_names = 0;
+ last = NULL;
+ name = nametable;
+ for(i=0;i<top;++i) {
+ if (last == NULL || !has_dupnames || strcmp(last+2,name+2)) {
+ ++num_names;
+ }
+ last = name;
+ name += entrysize;
+ }
+ tmp_vec = erts_alloc(ERTS_ALC_T_RE_TMP_BUF,
+ num_names * sizeof(Eterm));
+ /* Re-iterate and fill tmp_vec */
+ last = NULL;
+ name = nametable;
+ j = 0;
+ for(i=0;i<top;++i) {
+ if (last == NULL || !has_dupnames || strcmp(last+2,name+2)) {
+ tmp_vec[j++] = new_binary(BIF_P, (byte *) name+2, strlen(name+2));
+ }
+ last = name;
+ name += entrysize;
+ }
+ ASSERT(j == num_names);
+ hp = HAlloc(BIF_P, 3+2*j);
+ res = NIL;
+ for(i = j-1 ;i >= 0; --i) {
+ res = CONS(hp,tmp_vec[i],res);
+ hp += 2;
+ }
+ res = TUPLE2(hp,am_namelist,res);
+ erts_free_aligned_binary_bytes(temp_alloc);
+ erts_free(ERTS_ALC_T_RE_TMP_BUF, tmp_vec);
+ BIF_RET(res);
+
+ error:
+ /* tmp_vec never allocated when we reach here */
+ erts_free_aligned_binary_bytes(temp_alloc);
+ BIF_ERROR(BIF_P,BADARG);
+}
+
+
+
diff --git a/lib/stdlib/src/re.erl b/lib/stdlib/src/re.erl
index d8d529e6a4..4d6de1100d 100644
--- a/lib/stdlib/src/re.erl
+++ b/lib/stdlib/src/re.erl
@@ -28,11 +28,12 @@
| dotall | extended | firstline | multiline
| no_auto_capture | dupnames | ungreedy
| {newline, nl_spec()}| bsr_anycrlf
+ | no_start_optimize | ucp | never_utf
| bsr_unicode.
%%% BIFs
--export([compile/1, compile/2, run/2, run/3]).
+-export([compile/1, compile/2, run/2, run/3, inspect/2]).
-spec compile(Regexp) -> {ok, MP} | {error, ErrSpec} when
Regexp :: iodata(),
@@ -67,13 +68,13 @@ run(_, _) ->
Subject :: iodata() | unicode:charlist(),
RE :: mp() | iodata() | unicode:charlist(),
Options :: [Option],
- Option :: anchored | global | notbol | noteol | notempty
+ Option :: anchored | global | notbol | noteol | notempty | notempty_atstart
| {offset, non_neg_integer()} |
{newline, NLSpec :: nl_spec()} |
bsr_anycrlf | bsr_unicode | {capture, ValueSpec} |
{capture, ValueSpec, Type} | CompileOpt,
Type :: index | list | binary,
- ValueSpec :: all | all_but_first | first | none | ValueList,
+ ValueSpec :: all | all_but_first | all_names | first | none | ValueList,
ValueList :: [ValueID],
ValueID :: integer() | string() | atom(),
CompileOpt :: compile_option(),
@@ -88,6 +89,14 @@ run(_, _) ->
run(_, _, _) ->
erlang:nif_error(undef).
+-spec inspect(MP,Item) -> {namelist, [ binary() ]} when
+ MP :: mp(),
+ Item :: namelist.
+
+inspect(_,_) ->
+ erlang:nif_error(undef).
+
+
%%% End of BIFs
-spec split(Subject, RE) -> SplitList when
@@ -102,7 +111,7 @@ split(Subject,RE) ->
Subject :: iodata() | unicode:charlist(),
RE :: mp() | iodata() | unicode:charlist(),
Options :: [ Option ],
- Option :: anchored | notbol | noteol | notempty
+ Option :: anchored | notbol | noteol | notempty | notempty_atstart
| {offset, non_neg_integer()} | {newline, nl_spec()}
| bsr_anycrlf | bsr_unicode | {return, ReturnType}
| {parts, NumParts} | group | trim | CompileOpt,
@@ -295,7 +304,7 @@ replace(Subject,RE,Replacement) ->
RE :: mp() | iodata() | unicode:charlist(),
Replacement :: iodata() | unicode:charlist(),
Options :: [Option],
- Option :: anchored | global | notbol | noteol | notempty
+ Option :: anchored | global | notbol | noteol | notempty | notempty_atstart
| {offset, non_neg_integer()} | {newline, NLSpec} | bsr_anycrlf
| bsr_unicode | {return, ReturnType} | CompileOpt,
ReturnType :: iodata | list | binary,
@@ -509,7 +518,9 @@ check_for_crlf(_,L) ->
% SelectReturn = false | all | stirpfirst | none
% ConvertReturn = index | list | binary
% {capture, all} -> all (untouchded)
-% {capture, first} -> kept in argumentt list and Select all
+% {capture, all_names} -> if names are present: treated as a name {capture, [...]}
+% else: same as {capture, []}
+% {capture, first} -> kept in argument list and Select all
% {capture, all_but_first} -> removed from argument list and selects stripfirst
% {capture, none} -> removed from argument list and selects none
% {capture, []} -> removed from argument list and selects none
@@ -518,23 +529,30 @@ check_for_crlf(_,L) ->
% Call as process_parameters([],0,false,index,NeedClean)
-process_parameters([],InitialOffset, SelectReturn, ConvertReturn,_) ->
+process_parameters([],InitialOffset, SelectReturn, ConvertReturn,_,_) ->
{[], InitialOffset, SelectReturn, ConvertReturn};
-process_parameters([{offset, N} | T],_Init0,Select0,Return0,CC) ->
- process_parameters(T,N,Select0,Return0,CC);
-process_parameters([global | T],Init0,Select0,Return0,CC) ->
- process_parameters(T,Init0,Select0,Return0,CC);
-process_parameters([{capture,Values,Type}|T],Init0,Select0,_Return0,CC) ->
- process_parameters([{capture,Values}|T],Init0,Select0,Type,CC);
-process_parameters([{capture,Values}|T],Init0,Select0,Return0,CC) ->
+process_parameters([{offset, N} | T],_Init0,Select0,Return0,CC,RE) ->
+ process_parameters(T,N,Select0,Return0,CC,RE);
+process_parameters([global | T],Init0,Select0,Return0,CC,RE) ->
+ process_parameters(T,Init0,Select0,Return0,CC,RE);
+process_parameters([{capture,Values,Type}|T],Init0,Select0,_Return0,CC,RE) ->
+ process_parameters([{capture,Values}|T],Init0,Select0,Type,CC,RE);
+process_parameters([{capture,Values}|T],Init0,Select0,Return0,CC,RE) ->
% First process the rest to see if capture was already present
{NewTail, Init1, Select1, Return1} =
- process_parameters(T,Init0,Select0,Return0,CC),
+ process_parameters(T,Init0,Select0,Return0,CC,RE),
case Select1 of
false ->
case Values of
all ->
{[{capture,all} | NewTail], Init1, all, Return0};
+ all_names ->
+ case re:inspect(RE,namelist) of
+ {namelist, []} ->
+ {[{capture,first} | NewTail], Init1, none, Return0};
+ {namelist, List} ->
+ {[{capture,[0|List]} | NewTail], Init1, stripfirst, Return0}
+ end;
first ->
{[{capture,first} | NewTail], Init1, all, Return0};
all_but_first ->
@@ -553,20 +571,20 @@ process_parameters([{capture,Values}|T],Init0,Select0,Return0,CC) ->
% Found overriding further down list, ignore this one
{NewTail, Init1, Select1, Return1}
end;
-process_parameters([H|T],Init0,Select0,Return0,true) ->
+process_parameters([H|T],Init0,Select0,Return0,true,RE) ->
case copt(H) of
true ->
- process_parameters(T,Init0,Select0,Return0,true);
+ process_parameters(T,Init0,Select0,Return0,true,RE);
false ->
{NewT,Init,Select,Return} =
- process_parameters(T,Init0,Select0,Return0,true),
+ process_parameters(T,Init0,Select0,Return0,true,RE),
{[H|NewT],Init,Select,Return}
end;
-process_parameters([H|T],Init0,Select0,Return0,false) ->
+process_parameters([H|T],Init0,Select0,Return0,false,RE) ->
{NewT,Init,Select,Return} =
- process_parameters(T,Init0,Select0,Return0,false),
+ process_parameters(T,Init0,Select0,Return0,false,RE),
{[H|NewT],Init,Select,Return};
-process_parameters(_,_,_,_,_) ->
+process_parameters(_,_,_,_,_,_) ->
throw(badlist).
postprocess({match,[]},_,_,_,_) ->
@@ -723,7 +741,7 @@ do_grun(FlatSubject,Subject,Unicode,CRLF,RE,{Options0,NeedClean}) ->
{StrippedOptions, InitialOffset,
SelectReturn, ConvertReturn} =
case (catch
- process_parameters(Options0, 0, false, index, NeedClean)) of
+ process_parameters(Options0, 0, false, index, NeedClean,RE)) of
badlist ->
erlang:error(badarg,[Subject,RE,Options0]);
CorrectReturn ->
@@ -747,7 +765,7 @@ loopexec(Subject,RE,X,Y,Unicode,CRLF,Options) ->
loopexec(Subject,RE,A+B,Y,Unicode,CRLF,Options);
false ->
{match,M} =
- case re:run(Subject,RE,[{offset,X},notempty,
+ case re:run(Subject,RE,[{offset,X},notempty_atstart,
anchored]++Options) of
nomatch ->
{match,[]};
@@ -803,6 +821,12 @@ forward2(Chal,A,N,true,CRLF) ->
copt(caseless) ->
true;
+copt(no_start_optimize) ->
+ true;
+copt(never_utf) ->
+ true;
+copt(ucp) ->
+ true;
copt(dollar_endonly) ->
true;
copt(dotall) ->
@@ -833,6 +857,8 @@ copt(_) ->
runopt(notempty) ->
true;
+runopt(notempty_atstart) ->
+ true;
runopt(notbol) ->
true;
runopt(noteol) ->
diff --git a/lib/stdlib/test/re_SUITE.erl b/lib/stdlib/test/re_SUITE.erl
index d86e5f5b91..129f2b3e4c 100644
--- a/lib/stdlib/test/re_SUITE.erl
+++ b/lib/stdlib/test/re_SUITE.erl
@@ -25,7 +25,8 @@
split_autogen/1,split_options/1,split_specials/1,
error_handling/1,pcre_cve_2008_2371/1,
pcre_compile_workspace_overflow/1,re_infinite_loop/1,
- re_backwards_accented/1]).
+ re_backwards_accented/1,opt_dupnames/1,opt_all_names/1,inspect/1,
+ opt_no_start_optimize/1,opt_never_utf/1,opt_ucp/1]).
-include_lib("test_server/include/test_server.hrl").
-include_lib("kernel/include/file.hrl").
@@ -37,7 +38,9 @@ all() ->
replace_autogen, global_capture, replace_input_types,
replace_return, split_autogen, split_options,
split_specials, error_handling, pcre_cve_2008_2371,
- pcre_compile_workspace_overflow, re_infinite_loop, re_backwards_accented].
+ pcre_compile_workspace_overflow, re_infinite_loop,
+ re_backwards_accented, opt_dupnames, opt_all_names,
+ inspect, opt_no_start_optimize,opt_never_utf,opt_ucp].
groups() ->
[].
@@ -620,3 +623,155 @@ re_backwards_accented(Config) when is_list(Config) ->
[unicode,{capture,none}]),
?t:timetrap_cancel(Dog),
ok.
+opt_dupnames(doc) ->
+ "Check correct handling of dupnames option to re";
+opt_dupnames(Config) when is_list(Config) ->
+ Days = ["Monday","Tuesday","Wednesday","Thursday","Friday","Saturday","Sunday"],
+ _ = [ begin
+ Short = lists:sublist(Day,3),
+ {match,[Short]} =
+ re:run(Day,
+ "(?<DN>Mon|Fri|Sun)(?:day)?|(?<DN>Tue)(?:sday)?|"
+ "(?<DN>Wed)(?:nesday)?|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, ['DN'], list}])
+ end || Day <- Days ],
+ _ = [ begin
+ Short = list_to_binary(lists:sublist(Day,3)),
+ {match,[Short]} =
+ re:run(Day,
+ "(?<DN>Mon|Fri|Sun)(?:day)?|(?<DN>Tue)(?:sday)?|"
+ "(?<DN>Wed)(?:nesday)?|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, ['DN'], binary}])
+ end || Day <- Days ],
+ _ = [ begin
+ {match,[{0,3}]} =
+ re:run(Day,
+ "(?<DN>Mon|Fri|Sun)(?:day)?|(?<DN>Tue)(?:sday)?|"
+ "(?<DN>Wed)(?:nesday)?|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, ['DN'], index}])
+ end || Day <- Days ],
+ {match,[{0,1},{1,3},{7,1}]} = re:run("SMondayX","(?<Skrap>.)(?<DN>Mon|Fri|Sun)(?:day)?(?<Skrap2>.)|"
+ "(?<DN>Tue)(?:sday)?|(?<DN>Wed)nesday|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, ['Skrap','DN','Skrap2'],index}]),
+ {match,[{-1,0},{0,3},{-1,0}]} = re:run("Wednesday","(?<Skrap>.)(?<DN>Mon|Fri|Sun)(?:day)?(?<Skrap2>.)|"
+ "(?<DN>Tue)(?:sday)?|(?<DN>Wed)nesday|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, ['Skrap','DN','Skrap2'],index}]),
+ nomatch = re:run("Wednsday","(?<Skrap>.)(?<DN>Mon|Fri|Sun)(?:day)?(?<Skrap2>.)|"
+ "(?<DN>Tue)(?:sday)?|(?<DN>Wed)nesday|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, ['Skrap','DN','Skrap2'],index}]),
+ ok.
+
+opt_all_names(doc) ->
+ "Test capturing of all_names";
+opt_all_names(Config) when is_list(Config) ->
+ Days = ["Monday","Tuesday","Wednesday","Thursday","Friday","Saturday","Sunday"],
+ {match,[{1,3},{0,1},{7,1}]} = re:run("SMondayX","(?<Skrap>.)(?<DN>Mon|Fri|Sun)(?:day)?(?<Skrap2>.)|"
+ "(?<DN>Tue)(?:sday)?|(?<DN>Wed)nesday|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, all_names,index}]),
+ {match,[{0,3},{-1,0},{-1,0}]} = re:run("Wednesday","(?<Skrap>.)(?<DN>Mon|Fri|Sun)(?:day)?(?<Skrap2>.)|"
+ "(?<DN>Tue)(?:sday)?|(?<DN>Wed)nesday|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, all_names,index}]),
+
+ _ = [ begin
+ {match,[{0,3}]} =
+ re:run(Day,
+ "(?<DN>Mon|Fri|Sun)(?:day)?|(?<DN>Tue)(?:sday)?|"
+ "(?<DN>Wed)(?:nesday)?|(?<DN>Thu)(?:rsday)?|"
+ "(?<DN>Sat)(?:urday)?",
+ [dupnames, {capture, all_names, index}])
+ end || Day <- Days ],
+ _ = [ begin
+ match =
+ re:run(Day,
+ "(Mon|Fri|Sun)(?:day)?|(Tue)(?:sday)?|"
+ "(Wed)(?:nesday)?|(Thu)(?:rsday)?|"
+ "(Sat)(?:urday)?",
+ [dupnames, {capture, all_names, index}])
+ end || Day <- Days ],
+ {match,[{0,1},{-1,0},{-1,0}]} = re:run("A","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, index}]),
+ {match,[{-1,0},{0,1},{-1,0}]} = re:run("B","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, index}]),
+ {match,[{-1,0},{-1,0},{0,1}]} = re:run("C","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, index}]),
+ {match,[<<"A">>,<<>>,<<>>]} = re:run("A","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, binary}]),
+ {match,[<<>>,<<"B">>,<<>>]} = re:run("B","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, binary}]),
+ {match,[<<>>,<<>>,<<"C">>]} = re:run("C","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, binary}]),
+ {match,["A",[],[]]} = re:run("A","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, list}]),
+ {match,[[],"B",[]]} = re:run("B","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, list}]),
+ {match,[[],[],"C"]} = re:run("C","(?<A>A)|(?<B>B)|(?<C>C)",[{capture, all_names, list}]),
+ {match,[{-1,0},{-1,0},{0,1}]} = re:run("A","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, index}]),
+ {match,[{-1,0},{0,1},{-1,0}]} = re:run("B","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, index}]),
+ {match,[{0,1},{-1,0},{-1,0}]} = re:run("C","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, index}]),
+ {match,[<<>>,<<>>,<<"A">>]} = re:run("A","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, binary}]),
+ {match,[<<>>,<<"B">>,<<>>]} = re:run("B","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, binary}]),
+ {match,[<<"C">>,<<>>,<<>>]} = re:run("C","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, binary}]),
+ {match,[[],[],"A"]} = re:run("A","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, list}]),
+ {match,[[],"B",[]]} = re:run("B","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, list}]),
+ {match,["C",[],[]]} = re:run("C","(?<C>A)|(?<B>B)|(?<A>C)",[{capture, all_names, list}]),
+ {match,[[<<>>,<<>>,<<"C">>],
+ [<<>>,<<>>,<<"C">>],
+ [<<>>,<<>>,<<"C">>]]} = re:run("CCC","(?<A>A)|(?<B>B)|(?<C>C)",
+ [global,{capture, all_names, binary}]),
+ {match,[[<<"C">>,<<>>],
+ [<<>>,<<"B">>],
+ [<<"C">>,<<>>]]} = re:run("CBC","(?<A>A)|(?<B>B)|(?<A>C)",
+ [global,dupnames,{capture, all_names, binary}]),
+ {match,[[]]} = re:run("ABCE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,['A'],list}]),
+ {match,["D"]} = re:run("ABCDE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,['A'],list}]),
+ {match,["F"]} = re:run("ABCFE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,['A'],list}]),
+ {match,["F",[]]} = re:run("ABCFE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,['A','B'],list}]),
+ {match,[[],"E"]} = re:run("ABCE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,['A','B'],list}]),
+ {match,[[],"E"]} = re:run("ABCE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,all_names,list}]),
+ {match,[{-1,0},{3,1}]} = re:run("ABCE","(?<A>D)|(?<B>E)|(?<A>F)",[dupnames,{capture,all_names,index}]),
+ ok.
+
+inspect(doc) ->
+ "Test the minimal inspect function";
+inspect(Config) when is_list(Config)->
+ {ok,MP} = re:compile("(?<A>A)|(?<B>B)|(?<C>C)."),
+ {namelist,[<<"A">>,<<"B">>,<<"C">>]} = re:inspect(MP,namelist),
+ {ok,MPD} = re:compile("(?<A>A)|(?<B>B)|(?<A>C).",[dupnames]),
+ {namelist,[<<"A">>,<<"B">>]} = re:inspect(MPD,namelist),
+ {ok,MPN} = re:compile("(A)|(B)|(C)."),
+ {namelist,[]} = re:inspect(MPN,namelist),
+ {'EXIT',{badarg,_}} = (catch re:inspect(MPD,namelistk)),
+ {'EXIT',{badarg,_}} = (catch re:inspect({re_pattern,3,0,0,<<"kalle">>},namelist)),
+ ok.
+
+opt_no_start_optimize(doc) ->
+ "Test that the no_start_optimize compilation flag works";
+opt_no_start_optimize(Config) when is_list(Config) ->
+ {match, [{3,3}]} = re:run("DEFABC","(*COMMIT)ABC",[]), % Start optimization makes this result wrong!
+ nomatch = re:run("DEFABC","(*COMMIT)ABC",[no_start_optimize]), % This is the correct result...
+ ok.
+
+opt_never_utf(doc) ->
+ "Check that the never_utf option works";
+opt_never_utf(Config) when is_list(Config) ->
+ {match,[{0,3}]} = re:run("ABC","ABC",[never_utf]),
+ {match,[{0,3}]} = re:run("ABC","(*UTF)ABC",[]),
+ {ok,_} = re:compile("(*UTF)ABC"),
+ {ok,_} = re:compile("(*UTF)ABC",[unicode]),
+ {ok,_} = re:compile("(*UTF8)ABC"),
+ {'EXIT',{badarg,_}} = (catch re:run("ABC","ABC",[unicode,never_utf])),
+ {'EXIT',{badarg,_}} = (catch re:run("ABC","(*UTF)ABC",[never_utf])),
+ {'EXIT',{badarg,_}} = (catch re:run("ABC","(*UTF8)ABC",[never_utf])),
+ {error,_} = (catch re:compile("ABC",[unicode,never_utf])),
+ {error,_} = (catch re:compile("(*UTF)ABC",[never_utf])),
+ {error,_} = (catch re:compile("(*UTF8)ABC",[never_utf])),
+ ok.
+opt_ucp(doc) ->
+ "Check that the ucp option is passed to PCRE";
+opt_ucp(Config) when is_list(Config) ->
+ {match,[{0,1}]} = re:run([$a],"\\w",[unicode]),
+ {match,[{0,2}]} = re:run([229],"\\w",[unicode]), % Latin1 works without UCP, as we have a default
+ % Latin1 table
+ nomatch = re:run([1024],"\\w",[unicode]), % Latin1 word characters only, 1024 is not latin1
+ {match,[{0,2}]} = re:run([1024],"\\w",[unicode,ucp]), % Any Unicode word character works with 'ucp'
+ ok.