nokogumbo 1.3.0 → 1.5.0

data/gumbo-parser/src/string_buffer.c

@@ -26,11 +26,12 @@
 
 struct GumboInternalParser;
 
-static const size_t kDefaultStringBufferSize = 10;
+// Size chosen via statistical analysis of ~60K websites.
+// 99% of text nodes and 98% of attribute names/values fit in this initial size.
+static const size_t kDefaultStringBufferSize = 5;
 
-static void maybe_resize_string_buffer(
-    struct GumboInternalParser* parser, size_t additional_chars,
-    GumboStringBuffer* buffer) {
+static void maybe_resize_string_buffer(struct GumboInternalParser* parser,
+    size_t additional_chars, GumboStringBuffer* buffer) {
   size_t new_length = buffer->length + additional_chars;
   size_t new_capacity = buffer->capacity;
   while (new_capacity < new_length) {
@@ -52,9 +53,8 @@ void gumbo_string_buffer_init(
   output->capacity = kDefaultStringBufferSize;
 }
 
-void gumbo_string_buffer_reserve(
-    struct GumboInternalParser* parser, size_t min_capacity,
-    GumboStringBuffer* output) {
+void gumbo_string_buffer_reserve(struct GumboInternalParser* parser,
+    size_t min_capacity, GumboStringBuffer* output) {
   maybe_resize_string_buffer(parser, min_capacity - output->length, output);
 }
 
@@ -84,9 +84,8 @@ void gumbo_string_buffer_append_codepoint(
   }
 }
 
-void gumbo_string_buffer_append_string(
-    struct GumboInternalParser* parser, GumboStringPiece* str,
-    GumboStringBuffer* output) {
+void gumbo_string_buffer_append_string(struct GumboInternalParser* parser,
+    GumboStringPiece* str, GumboStringBuffer* output) {
   maybe_resize_string_buffer(parser, str->length, output);
   memcpy(output->data + output->length, str->data, str->length);
   output->length += str->length;
@@ -100,6 +99,11 @@ char* gumbo_string_buffer_to_string(
   return buffer;
 }
 
+void gumbo_string_buffer_clear(
+    struct GumboInternalParser* parser, GumboStringBuffer* input) {
+  input->length = 0;
+}
+
 void gumbo_string_buffer_destroy(
     struct GumboInternalParser* parser, GumboStringBuffer* buffer) {
   gumbo_parser_deallocate(parser, buffer->data);

data/gumbo-parser/src/string_buffer.h

@@ -51,9 +51,8 @@ void gumbo_string_buffer_init(
 // Ensures that the buffer contains at least a certain amount of space.  Most
 // useful with snprintf and the other length-delimited string functions, which
 // may want to write directly into the buffer.
-void gumbo_string_buffer_reserve(
-    struct GumboInternalParser* parser, size_t min_capacity,
-    GumboStringBuffer* output);
+void gumbo_string_buffer_reserve(struct GumboInternalParser* parser,
+    size_t min_capacity, GumboStringBuffer* output);
 
 // Appends a single Unicode codepoint onto the end of the GumboStringBuffer.
 // This is essentially a UTF-8 encoder, and may add 1-4 bytes depending on the
@@ -62,14 +61,18 @@ void gumbo_string_buffer_append_codepoint(
     struct GumboInternalParser* parser, int c, GumboStringBuffer* output);
 
 // Appends a string onto the end of the GumboStringBuffer.
-void gumbo_string_buffer_append_string(
-    struct GumboInternalParser* parser, GumboStringPiece* str,
-    GumboStringBuffer* output);
+void gumbo_string_buffer_append_string(struct GumboInternalParser* parser,
+    GumboStringPiece* str, GumboStringBuffer* output);
 
 // Converts this string buffer to const char*, alloctaing a new buffer for it.
 char* gumbo_string_buffer_to_string(
     struct GumboInternalParser* parser, GumboStringBuffer* input);
 
+// Reinitialize this string buffer.  This clears it by setting length=0.  It
+// does not zero out the buffer itself.
+void gumbo_string_buffer_clear(
+    struct GumboInternalParser* parser, GumboStringBuffer* input);
+
 // Deallocates this GumboStringBuffer.
 void gumbo_string_buffer_destroy(
     struct GumboInternalParser* parser, GumboStringBuffer* buffer);

data/gumbo-parser/src/string_piece.c

@@ -25,23 +25,22 @@
 
 struct GumboInternalParser;
 
-const GumboStringPiece kGumboEmptyString = { NULL, 0 };
+const GumboStringPiece kGumboEmptyString = {NULL, 0};
 
 bool gumbo_string_equals(
     const GumboStringPiece* str1, const GumboStringPiece* str2) {
   return str1->length == str2->length &&
-      !memcmp(str1->data, str2->data, str1->length);
+         !memcmp(str1->data, str2->data, str1->length);
 }
 
 bool gumbo_string_equals_ignore_case(
     const GumboStringPiece* str1, const GumboStringPiece* str2) {
   return str1->length == str2->length &&
-      !strncasecmp(str1->data, str2->data, str1->length);
+         !strncasecmp(str1->data, str2->data, str1->length);
 }
 
-void gumbo_string_copy(
-    struct GumboInternalParser* parser, GumboStringPiece* dest,
-    const GumboStringPiece* source) {
+void gumbo_string_copy(struct GumboInternalParser* parser,
+    GumboStringPiece* dest, const GumboStringPiece* source) {
   dest->length = source->length;
   char* buffer = gumbo_parser_allocate(parser, source->length);
   memcpy(buffer, source->data, source->length);

data/gumbo-parser/src/string_piece.h

@@ -28,9 +28,8 @@ struct GumboInternalParser;
 // Performs a deep-copy of an GumboStringPiece, allocating a fresh buffer in the
 // destination and copying over the characters from source.  Dest should be
 // empty, with no buffer allocated; otherwise, this leaks it.
-void gumbo_string_copy(
-    struct GumboInternalParser* parser, GumboStringPiece* dest,
-    const GumboStringPiece* source);
+void gumbo_string_copy(struct GumboInternalParser* parser,
+    GumboStringPiece* dest, const GumboStringPiece* source);
 
 #ifdef __cplusplus
 }

data/gumbo-parser/src/tag.c

@@ -18,164 +18,18 @@
 
 #include <assert.h>
 #include <ctype.h>
-#include <strings.h>    // For strcasecmp.
+#include <string.h>
 
-// NOTE(jdtang): Keep this in sync with the GumboTag enum in the header.
-// TODO(jdtang): Investigate whether there're efficiency benefits to putting the
-// most common tag names first, or to putting them in alphabetical order and
-// using a binary search.
 const char* kGumboTagNames[] = {
-  "html",
-  "head",
-  "title",
-  "base",
-  "link",
-  "meta",
-  "style",
-  "script",
-  "noscript",
-  "template",
-  "body",
-  "article",
-  "section",
-  "nav",
-  "aside",
-  "h1",
-  "h2",
-  "h3",
-  "h4",
-  "h5",
-  "h6",
-  "hgroup",
-  "header",
-  "footer",
-  "address",
-  "p",
-  "hr",
-  "pre",
-  "blockquote",
-  "ol",
-  "ul",
-  "li",
-  "dl",
-  "dt",
-  "dd",
-  "figure",
-  "figcaption",
-  "main",
-  "div",
-  "a",
-  "em",
-  "strong",
-  "small",
-  "s",
-  "cite",
-  "q",
-  "dfn",
-  "abbr",
-  "data",
-  "time",
-  "code",
-  "var",
-  "samp",
-  "kbd",
-  "sub",
-  "sup",
-  "i",
-  "b",
-  "u",
-  "mark",
-  "ruby",
-  "rt",
-  "rp",
-  "bdi",
-  "bdo",
-  "span",
-  "br",
-  "wbr",
-  "ins",
-  "del",
-  "image",
-  "img",
-  "iframe",
-  "embed",
-  "object",
-  "param",
-  "video",
-  "audio",
-  "source",
-  "track",
-  "canvas",
-  "map",
-  "area",
-  "math",
-  "mi",
-  "mo",
-  "mn",
-  "ms",
-  "mtext",
-  "mglyph",
-  "malignmark",
-  "annotation-xml",
-  "svg",
-  "foreignobject",
-  "desc",
-  "table",
-  "caption",
-  "colgroup",
-  "col",
-  "tbody",
-  "thead",
-  "tfoot",
-  "tr",
-  "td",
-  "th",
-  "form",
-  "fieldset",
-  "legend",
-  "label",
-  "input",
-  "button",
-  "select",
-  "datalist",
-  "optgroup",
-  "option",
-  "textarea",
-  "keygen",
-  "output",
-  "progress",
-  "meter",
-  "details",
-  "summary",
-  "menu",
-  "menuitem",
-  "applet",
-  "acronym",
-  "bgsound",
-  "dir",
-  "frame",
-  "frameset",
-  "noframes",
-  "isindex",
-  "listing",
-  "xmp",
-  "nextid",
-  "noembed",
-  "plaintext",
-  "rb",
-  "strike",
-  "basefont",
-  "big",
-  "blink",
-  "center",
-  "font",
-  "marquee",
-  "multicol",
-  "nobr",
-  "spacer",
-  "tt",
-  "",                   // TAG_UNKNOWN
-  "",                   // TAG_LAST
+#include "tag_strings.h"
+    "",  // TAG_UNKNOWN
+    "",  // TAG_LAST
+};
+
+static const unsigned char kGumboTagSizes[] = {
+#include "tag_sizes.h"
+    0,  // TAG_UNKNOWN
+    0,  // TAG_LAST
 };
 
 const char* gumbo_normalized_tagname(GumboTag tag) {
@@ -183,7 +37,6 @@ const char* gumbo_normalized_tagname(GumboTag tag) {
   return kGumboTagNames[tag];
 }
 
-// TODO(jdtang): Add test for this.
 void gumbo_tag_from_original_text(GumboStringPiece* text) {
   if (text->data == NULL) {
     return;
@@ -195,11 +48,11 @@ void gumbo_tag_from_original_text(GumboStringPiece* text) {
   if (text->data[1] == '/') {
     // End tag.
     assert(text->length >= 3);
-    text->data += 2;    // Move past </
+    text->data += 2;  // Move past </
     text->length -= 3;
   } else {
     // Start tag.
-    text->data += 1;    // Move past <
+    text->data += 1;  // Move past <
     text->length -= 2;
     // strnchr is apparently not a standard C library function, so I loop
     // explicitly looking for whitespace or other illegal tag characters.
@@ -212,14 +65,31 @@ void gumbo_tag_from_original_text(GumboStringPiece* text) {
   }
 }
 
-GumboTag gumbo_tag_enum(const char* tagname) {
-  for (int i = 0; i < GUMBO_TAG_LAST; ++i) {
-    // TODO(jdtang): strcasecmp is non-portable, so if we want to support
-    // non-GCC compilers, we'll need some #ifdef magic.  This source already has
-    // pretty significant issues with MSVC6 anyway.
-    if (strcasecmp(tagname, kGumboTagNames[i]) == 0) {
-      return i;
+static int case_memcmp(const char* s1, const char* s2, unsigned int n) {
+  while (n--) {
+    unsigned char c1 = tolower(*s1++);
+    unsigned char c2 = tolower(*s2++);
+    if (c1 != c2) return (int) c1 - (int) c2;
+  }
+  return 0;
+}
+
+#include "tag_gperf.h"
+#define TAG_MAP_SIZE (sizeof(kGumboTagMap) / sizeof(kGumboTagMap[0]))
+
+GumboTag gumbo_tagn_enum(const char* tagname, unsigned int length) {
+  if (length) {
+    unsigned int key = tag_hash(tagname, length);
+    if (key < TAG_MAP_SIZE) {
+      GumboTag tag = kGumboTagMap[key];
+      if (length == kGumboTagSizes[(int) tag] &&
+          !case_memcmp(tagname, kGumboTagNames[(int) tag], length))
+        return tag;
     }
   }
   return GUMBO_TAG_UNKNOWN;
 }
+
+GumboTag gumbo_tag_enum(const char* tagname) {
+  return gumbo_tagn_enum(tagname, strlen(tagname));
+}

data/gumbo-parser/src/tag.in

@@ -0,0 +1,150 @@
+html
+head
+title
+base
+link
+meta
+style
+script
+noscript
+template
+body
+article
+section
+nav
+aside
+h1
+h2
+h3
+h4
+h5
+h6
+hgroup
+header
+footer
+address
+p
+hr
+pre
+blockquote
+ol
+ul
+li
+dl
+dt
+dd
+figure
+figcaption
+main
+div
+a
+em
+strong
+small
+s
+cite
+q
+dfn
+abbr
+data
+time
+code
+var
+samp
+kbd
+sub
+sup
+i
+b
+u
+mark
+ruby
+rt
+rp
+bdi
+bdo
+span
+br
+wbr
+ins
+del
+image
+img
+iframe
+embed
+object
+param
+video
+audio
+source
+track
+canvas
+map
+area
+math
+mi
+mo
+mn
+ms
+mtext
+mglyph
+malignmark
+annotation-xml
+svg
+foreignobject
+desc
+table
+caption
+colgroup
+col
+tbody
+thead
+tfoot
+tr
+td
+th
+form
+fieldset
+legend
+label
+input
+button
+select
+datalist
+optgroup
+option
+textarea
+keygen
+output
+progress
+meter
+details
+summary
+menu
+menuitem
+applet
+acronym
+bgsound
+dir
+frame
+frameset
+noframes
+isindex
+listing
+xmp
+nextid
+noembed
+plaintext
+rb
+strike
+basefont
+big
+blink
+center
+font
+marquee
+multicol
+nobr
+spacer
+tt
+rtc