X-Git-Url: https://git.wh0rd.org/?a=blobdiff_plain;f=src%2Ffcformat.c;h=d8518f458252fba1abc6ee4b54111807c750a7c7;hb=HEAD;hp=44730094c85a788b030cac1a427b4a75ccaf453c;hpb=85c7fb67ce9f77574f71de7d9b69867bb974cd48;p=fontconfig.git

diff --git a/src/fcformat.c b/src/fcformat.c
index 4473009..d8518f4 100644
--- a/src/fcformat.c
+++ b/src/fcformat.c
@@ -7,15 +7,15 @@
  * documentation for any purpose is hereby granted without fee, provided that
  * the above copyright notice appear in all copies and that both that
  * copyright notice and this permission notice appear in supporting
- * documentation, and that the name of Keith Packard not be used in
+ * documentation, and that the name of the author(s) not be used in
  * advertising or publicity pertaining to distribution of the software without
- * specific, written prior permission.  Keith Packard makes no
+ * specific, written prior permission.  The authors make no
  * representations about the suitability of this software for any purpose.  It
  * is provided "as is" without express or implied warranty.
  *
- * KEITH PACKARD DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
+ * THE AUTHOR(S) DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE,
  * INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN NO
- * EVENT SHALL KEITH PACKARD BE LIABLE FOR ANY SPECIAL, INDIRECT OR
+ * EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY SPECIAL, INDIRECT OR
  * CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
  * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
  * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
@@ -28,14 +28,59 @@
 #include <stdarg.h>
 
 
-/*
+/* The language is documented in doc/fcformat.fncs
+ * These are the features implemented:
+ *
+ * simple	%{elt}
+ * width	%width{elt}
+ * index	%{elt[idx]}
+ * name=	%{elt=}
+ * :name=	%{:elt}
+ * default	%{elt:-word}
+ * count	%{#elt}
+ * subexpr	%{{expr}}
+ * filter-out	%{-elt1,elt2,elt3{expr}}
+ * filter-in	%{+elt1,elt2,elt3{expr}}
+ * conditional	%{?elt1,elt2,!elt3{}{}}
+ * enumerate	%{[]elt1,elt2{expr}}
+ * langset	langset enumeration using the same syntax
+ * builtin	%{=blt}
+ * convert	%{elt|conv1|conv2|conv3}
+ *
+ * converters:
+ * basename	FcStrBasename
+ * dirname	FcStrDirname
+ * downcase	FcStrDowncase
+ * shescape
+ * cescape
+ * xmlescape
+ * delete	delete chars
+ * escape	escape chars
+ * translate	translate chars
+ *
+ * builtins:
+ * unparse	FcNameUnparse
+ * fcmatch	fc-match default
+ * fclist	fc-list default
+ * fccat	fc-cat default
+ * pkgkit	PackageKit package tag format
+ *
+ *
  * Some ideas for future syntax extensions:
  *
+ * - verbose builtin that is like FcPatternPrint
  * - allow indexing subexprs using '%{[idx]elt1,elt2{subexpr}}'
- * - allow indexing simple tags using '%{elt[idx]}'
+ * - allow indexing in +, -, ? filtering?
  * - conditional/filtering/deletion on binding (using '(w)'/'(s)'/'(=)' notation)
  */
 
+
+#define FCCAT_FORMAT	"\"%{file|basename|cescape}\" %{index} \"%{-file{%{=unparse|cescape}}}\""
+#define FCMATCH_FORMAT	"%{file:-<unknown filename>|basename}: \"%{family[0]:-<unknown family>}\" \"%{style[0]:-<unknown style>}\""
+#define FCLIST_FORMAT	"%{?file{%{file}: }}%{-file{%{=unparse}}}"
+#define PKGKIT_FORMAT	"%{[]family{font(%{family|downcase|delete( )})\n}}%{[]lang{font(:lang=%{lang|downcase|translate(_,-)})\n}}"
+
+
 static void
 message (const char *fmt, ...)
 {
@@ -216,30 +261,64 @@ read_chars (FcFormatContext *c,
     return FcTrue;
 }
 
+static FcBool
+FcPatternFormatToBuf (FcPattern     *pat,
+		      const FcChar8 *format,
+		      FcStrBuf      *buf);
+
 static FcBool
 interpret_builtin (FcFormatContext *c,
 		   FcPattern       *pat,
 		   FcStrBuf        *buf)
 {
-    if (!expect_char (c, '='))
-	return FcFalse;
+    FcChar8       *new_str;
+    FcBool         ret;
 
-    if (!read_word (c))
+    if (!expect_char (c, '=') ||
+	!read_word (c))
 	return FcFalse;
+
+    /* try simple builtins first */
+    if (0) { }
 #define BUILTIN(name, func) \
     else if (0 == strcmp ((const char *) c->word, name))\
-	return func (c, pat, buf)
-#if 0
-    BUILTIN  ("unparse",  FcNameUnparse);
-    BUILTIN  ("verbose",  FcPatternPrint);
-    BUILTIN2 ("fcmatch",  FcStrDirname);
-    BUILTIN2 ("fclist",   FcStrDirname);
-    BUILTIN2 ("pkgkit",   FcStrDirname);
-#endif
-
-    message ("unknown builtin \"%s\"",
-	     c->word);
-    return FcFalse;
+	do { new_str = func (pat); ret = FcTrue; } while (0)
+    BUILTIN ("unparse",  FcNameUnparse);
+ /* BUILTIN ("verbose",  FcPatternPrint); XXX */
+#undef BUILTIN
+    else
+	ret = FcFalse;
+
+    if (ret)
+    {
+	if (new_str)
+	{
+	    FcStrBufString (buf, new_str);
+	    free (new_str);
+	    return FcTrue;
+	}
+	else
+	    return FcFalse;
+    }
+
+    /* now try our custom formats */
+    if (0) { }
+#define BUILTIN(name, format) \
+    else if (0 == strcmp ((const char *) c->word, name))\
+	ret = FcPatternFormatToBuf (pat, (const FcChar8 *) format, buf)
+    BUILTIN ("fccat",    FCCAT_FORMAT);
+    BUILTIN ("fcmatch",  FCMATCH_FORMAT);
+    BUILTIN ("fclist",   FCLIST_FORMAT);
+    BUILTIN ("pkgkit",   PKGKIT_FORMAT);
+#undef BUILTIN
+    else
+	ret = FcFalse;
+
+    if (!ret)
+	message ("unknown builtin \"%s\"",
+		 c->word);
+
+    return ret;
 }
 
 static FcBool
@@ -345,9 +424,9 @@ maybe_skip_subexpr (FcFormatContext *c)
 }
 
 static FcBool
-interpret_filter (FcFormatContext *c,
-		  FcPattern       *pat,
-		  FcStrBuf        *buf)
+interpret_filter_in (FcFormatContext *c,
+		     FcPattern       *pat,
+		     FcStrBuf        *buf)
 {
     FcObjectSet  *os;
     FcPattern    *subpat;
@@ -361,6 +440,7 @@ interpret_filter (FcFormatContext *c,
 
     do
     {
+	/* XXX binding */
 	if (!read_word (c) ||
 	    !FcObjectSetAdd (os, (const char *) c->word))
 	{
@@ -382,9 +462,9 @@ interpret_filter (FcFormatContext *c,
 }
 
 static FcBool
-interpret_delete (FcFormatContext *c,
-		  FcPattern       *pat,
-		  FcStrBuf        *buf)
+interpret_filter_out (FcFormatContext *c,
+		      FcPattern       *pat,
+		      FcStrBuf        *buf)
 {
     FcPattern    *subpat;
 
@@ -438,9 +518,8 @@ interpret_cond (FcFormatContext *c,
 
 	pass = pass &&
 	       (negate ^
-		(FcResultMatch == FcPatternGet (pat,
-						(const char *) c->word,
-						0, &v)));
+		(FcResultMatch ==
+		 FcPatternGet (pat, (const char *) c->word, 0, &v)));
     }
     while (consume_char (c, ','));
 
@@ -494,6 +573,124 @@ interpret_count (FcFormatContext *c,
     return FcTrue;
 }
 
+static FcBool
+interpret_enumerate (FcFormatContext *c,
+		     FcPattern       *pat,
+		     FcStrBuf        *buf)
+{
+    FcObjectSet   *os;
+    FcPattern     *subpat;
+    const FcChar8 *format_save;
+    int            idx;
+    FcBool         ret, done;
+    FcStrList      *lang_strs;
+
+    if (!expect_char (c, '[') ||
+	!expect_char (c, ']'))
+	return FcFalse;
+
+    os = FcObjectSetCreate ();
+    if (!os)
+	return FcFalse;
+
+    ret = FcTrue;
+
+    do
+    {
+	if (!read_word (c) ||
+	    !FcObjectSetAdd (os, (const char *) c->word))
+	{
+	    FcObjectSetDestroy (os);
+	    return FcFalse;
+	}
+    }
+    while (consume_char (c, ','));
+
+    /* If we have one element and it's of type FcLangSet, we want
+     * to enumerate the languages in it. */
+    lang_strs = NULL;
+    if (os->nobject == 1)
+    {
+	FcLangSet *langset;
+	if (FcResultMatch ==
+	    FcPatternGetLangSet (pat, os->objects[0], 0, &langset))
+	{
+	    FcStrSet *ss;
+	    if (!(ss = FcLangSetGetLangs (langset)) ||
+		!(lang_strs = FcStrListCreate (ss)))
+		goto bail0;
+	}
+    }
+
+    subpat = FcPatternDuplicate (pat);
+    if (!subpat)
+	goto bail0;
+
+    format_save = c->format;
+    idx = 0;
+    do
+    {
+	int i;
+
+	done = FcTrue;
+
+	if (lang_strs)
+	{
+	    FcChar8 *lang;
+
+	    FcPatternDel (subpat, os->objects[0]);
+	    if ((lang = FcStrListNext (lang_strs)))
+	    {
+		/* XXX binding? */
+		FcPatternAddString (subpat, os->objects[0], lang);
+		done = FcFalse;
+	    }
+	}
+	else
+	{
+	    for (i = 0; i < os->nobject; i++)
+	    {
+		FcValue v;
+
+		/* XXX this can be optimized by accessing valuelist linked lists
+		 * directly and remembering where we were.  Most (all) value lists
+		 * in normal uses are pretty short though (language tags are
+		 * stored as a LangSet, not separate values.). */
+		FcPatternDel (subpat, os->objects[i]);
+		if (FcResultMatch ==
+		    FcPatternGet (pat, os->objects[i], idx, &v))
+		{
+		    /* XXX binding */
+		    FcPatternAdd (subpat, os->objects[i], v, FcFalse);
+		    done = FcFalse;
+		}
+	    }
+	}
+
+	if (!done)
+	{
+	    c->format = format_save;
+	    ret = interpret_subexpr (c, subpat, buf);
+	    if (!ret)
+		goto bail;
+	}
+
+	idx++;
+    } while (!done);
+
+    if (c->format == format_save)
+	skip_subexpr (c);
+
+bail:
+    FcPatternDestroy (subpat);
+bail0:
+    if (lang_strs)
+	FcStrListDone (lang_strs);
+    FcObjectSetDestroy (os);
+
+    return ret;
+}
+
 static FcBool
 interpret_simple (FcFormatContext *c,
 		  FcPattern       *pat,
@@ -502,6 +699,8 @@ interpret_simple (FcFormatContext *c,
     FcPatternElt *e;
     FcBool        add_colon = FcFalse;
     FcBool        add_elt_name = FcFalse;
+    int           idx;
+    FcChar8      *else_string;
 
     if (consume_char (c, ':'))
 	add_colon = FcTrue;
@@ -509,14 +708,47 @@ interpret_simple (FcFormatContext *c,
     if (!read_word (c))
 	return FcFalse;
 
+    idx = -1;
+    if (consume_char (c, '['))
+    {
+	idx = strtol ((const char *) c->format, (char **) &c->format, 10);
+	if (idx < 0)
+	{
+	    message ("expected non-negative number at %d",
+		     c->format-1 - c->format_orig + 1);
+	    return FcFalse;
+	}
+	if (!expect_char (c, ']'))
+	    return FcFalse;
+    }
+
     if (consume_char (c, '='))
 	add_elt_name = FcTrue;
 
+    /* modifiers */
+    else_string = NULL;
+    if (consume_char (c, ':'))
+    {
+	FcChar8 *orig;
+	/* divert the c->word for now */
+	orig = c->word;
+	c->word = c->word + strlen ((const char *) c->word) + 1;
+	/* for now we just support 'default value' */
+	if (!expect_char (c, '-') ||
+	    !read_chars (c, '|'))
+	{
+	    c->word = orig;
+	    return FcFalse;
+	}
+	else_string = c->word;
+	c->word = orig;
+    }
+
     e = FcPatternObjectFindElt (pat,
 				FcObjectFromName ((const char *) c->word));
-    if (e)
+    if (e || else_string)
     {
-	FcValueListPtr l;
+	FcValueListPtr l = NULL;
 
 	if (add_colon)
 	    FcStrBufChar (buf, ':');
@@ -526,8 +758,33 @@ interpret_simple (FcFormatContext *c,
 	    FcStrBufChar (buf, '=');
 	}
 
-	l = FcPatternEltValues(e);
-	FcNameUnparseValueList (buf, l, '\0');
+	if (e)
+	    l = FcPatternEltValues(e);
+
+	if (idx != -1)
+	{
+	    while (l && idx > 0)
+	    {
+		l = FcValueListNext(l);
+		idx--;
+	    }
+	    if (l && idx == 0)
+	    {
+		if (!FcNameUnparseValue (buf, &l->value, '\0'))
+		    return FcFalse;
+	    }
+	    else goto notfound;
+        }
+	else if (l)
+	{
+	    FcNameUnparseValueList (buf, l, '\0');
+	}
+	else
+	{
+    notfound:
+	    if (else_string)
+		FcStrBufString (buf, else_string);
+	}
     }
 
     return FcTrue;
@@ -535,9 +792,11 @@ interpret_simple (FcFormatContext *c,
 
 static FcBool
 cescape (FcFormatContext *c,
-	 FcStrBuf        *buf,
-	 const FcChar8   *str)
+	 const FcChar8   *str,
+	 FcStrBuf        *buf)
 {
+    /* XXX escape \n etc? */
+
     while(*str)
     {
 	switch (*str)
@@ -554,8 +813,8 @@ cescape (FcFormatContext *c,
 
 static FcBool
 shescape (FcFormatContext *c,
-	  FcStrBuf        *buf,
-	  const FcChar8   *str)
+	  const FcChar8   *str,
+	  FcStrBuf        *buf)
 {
     FcStrBufChar (buf, '\'');
     while(*str)
@@ -572,9 +831,11 @@ shescape (FcFormatContext *c,
 
 static FcBool
 xmlescape (FcFormatContext *c,
-	   FcStrBuf        *buf,
-	   const FcChar8   *str)
+	   const FcChar8   *str,
+	   FcStrBuf        *buf)
 {
+    /* XXX escape \n etc? */
+
     while(*str)
     {
 	switch (*str)
@@ -591,8 +852,8 @@ xmlescape (FcFormatContext *c,
 
 static FcBool
 delete_chars (FcFormatContext *c,
-	      FcStrBuf        *buf,
-	      const FcChar8   *str)
+	      const FcChar8   *str,
+	      FcStrBuf        *buf)
 {
     /* XXX not UTF-8 aware */
 
@@ -624,8 +885,8 @@ delete_chars (FcFormatContext *c,
 
 static FcBool
 escape_chars (FcFormatContext *c,
-	      FcStrBuf        *buf,
-	      const FcChar8   *str)
+	      const FcChar8   *str,
+	      FcStrBuf        *buf)
 {
     /* XXX not UTF-8 aware */
 
@@ -659,8 +920,8 @@ escape_chars (FcFormatContext *c,
 
 static FcBool
 translate_chars (FcFormatContext *c,
-		 FcStrBuf        *buf,
-		 const FcChar8   *str)
+		 const FcChar8   *str,
+		 FcStrBuf        *buf)
 {
     char *from, *to, repeat;
     int from_len, to_len;
@@ -676,7 +937,7 @@ translate_chars (FcFormatContext *c,
     from_len = strlen (from);
     to = from + from_len + 1;
 
-    /* hack: we temporarily diverge c->word */
+    /* hack: we temporarily divert c->word */
     c->word = (FcChar8 *) to;
     if (!read_chars (c, ')'))
     {
@@ -726,19 +987,17 @@ interpret_convert (FcFormatContext *c,
     FcChar8        buf_static[8192];
     FcBool         ret;
 
-    if (!expect_char (c, '|'))
+    if (!expect_char (c, '|') ||
+	!read_word (c))
 	return FcFalse;
 
-    /* nul-terminate the buffer */
+    /* prepare the buffer */
     FcStrBufChar (buf, '\0');
     if (buf->failed)
 	return FcFalse;
     str = buf->buf + start;
     buf->len = start;
 
-    if (!read_word (c))
-	return FcFalse;
-
     /* try simple converters first */
     if (0) { }
 #define CONVERTER(name, func) \
@@ -755,7 +1014,6 @@ interpret_convert (FcFormatContext *c,
     {
 	if (new_str)
 	{
-	    /* replace in the buffer */
 	    FcStrBufString (buf, new_str);
 	    free (new_str);
 	    return FcTrue;
@@ -770,7 +1028,7 @@ interpret_convert (FcFormatContext *c,
     if (0) { }
 #define CONVERTER(name, func) \
     else if (0 == strcmp ((const char *) c->word, name))\
-	ret = func (c, &new_buf, str)
+	ret = func (c, str, &new_buf)
     CONVERTER ("cescape",   cescape);
     CONVERTER ("shescape",  shescape);
     CONVERTER ("xmlescape", xmlescape);
@@ -870,13 +1128,14 @@ interpret_percent (FcFormatContext *c,
     start = buf->len;
 
     switch (*c->format) {
-    case '=': ret = interpret_builtin (c, pat, buf); break;
-    case '{': ret = interpret_subexpr (c, pat, buf); break;
-    case '+': ret = interpret_filter  (c, pat, buf); break;
-    case '-': ret = interpret_delete  (c, pat, buf); break;
-    case '?': ret = interpret_cond    (c, pat, buf); break;
-    case '#': ret = interpret_count   (c, pat, buf); break;
-    default:  ret = interpret_simple  (c, pat, buf); break;
+    case '=': ret = interpret_builtin    (c, pat, buf); break;
+    case '{': ret = interpret_subexpr    (c, pat, buf); break;
+    case '+': ret = interpret_filter_in  (c, pat, buf); break;
+    case '-': ret = interpret_filter_out (c, pat, buf); break;
+    case '?': ret = interpret_cond       (c, pat, buf); break;
+    case '#': ret = interpret_count      (c, pat, buf); break;
+    case '[': ret = interpret_enumerate  (c, pat, buf); break;
+    default:  ret = interpret_simple     (c, pat, buf); break;
     }
 
     return ret &&
@@ -910,22 +1169,37 @@ interpret_expr (FcFormatContext *c,
     return FcTrue;
 }
 
+static FcBool
+FcPatternFormatToBuf (FcPattern     *pat,
+		      const FcChar8 *format,
+		      FcStrBuf      *buf)
+{
+    FcFormatContext c;
+    FcChar8         word_static[1024];
+    FcBool          ret;
+
+    if (!FcFormatContextInit (&c, format, word_static, sizeof (word_static)))
+	return FcFalse;
+
+    ret = interpret_expr (&c, pat, buf, '\0');
+
+    FcFormatContextDone (&c);
+
+    return ret;
+}
+
 FcChar8 *
-FcPatternFormat (FcPattern *pat, const FcChar8 *format)
+FcPatternFormat (FcPattern *pat,
+		 const FcChar8 *format)
 {
     FcStrBuf        buf;
-    FcChar8         word_static[1024];
     FcChar8         buf_static[8192 - 1024];
-    FcFormatContext c;
     FcBool          ret;
 
     FcStrBufInit (&buf, buf_static, sizeof (buf_static));
-    if (!FcFormatContextInit (&c, format, word_static, sizeof (word_static)))
-	return NULL;
 
-    ret = interpret_expr (&c, pat, &buf, '\0');
+    ret = FcPatternFormatToBuf (pat, format, &buf);
 
-    FcFormatContextDone (&c);
     if (ret)
 	return FcStrBufDone (&buf);
     else