lib/libxml2
1
0
Fork 0
mirror of https://gitlab.gnome.org/GNOME/libxml2.git synced 2025-10-24 13:33:01 +03:00
Code Activity

Convert of the HTML parser to new input buffers

Browse Source 
Changes similar to the ones done in the XML parser for the
routines which are not shared.
This commit is contained in:
Daniel Veillard
2012-07-16 14:56:50 +08:00
parent dbf5411b21
commit a78d803639
1 changed files with 29 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

61
HTMLparser.c
View File

@@ -44,6 +44,9 @@
#include <libxml/globals.h> #include <libxml/globals.h>
#include <libxml/uri.h> #include <libxml/uri.h>
#include "buf.h"
#include "enc.h"
#define HTML_MAX_NAMELEN 1000 #define HTML_MAX_NAMELEN 1000
#define HTML_PARSER_BIG_BUFFER_SIZE 1000 #define HTML_PARSER_BIG_BUFFER_SIZE 1000
#define HTML_PARSER_BUFFER_SIZE 100 #define HTML_PARSER_BUFFER_SIZE 100
@@ -3509,19 +3512,16 @@ htmlCheckEncodingDirect(htmlParserCtxtPtr ctxt, const xmlChar *encoding) {
* convert as much as possible to the parser reading buffer. * convert as much as possible to the parser reading buffer.
*/ */
processed = ctxt->input->cur - ctxt->input->base; processed = ctxt->input->cur - ctxt->input->base;
xmlBufferShrink(ctxt->input->buf->buffer, processed); xmlBufShrink(ctxt->input->buf->buffer, processed);
nbchars = xmlCharEncInFunc(ctxt->input->buf->encoder, nbchars = xmlCharEncInput(ctxt->input->buf);
ctxt->input->buf->buffer,
ctxt->input->buf->raw);
if (nbchars < 0) { if (nbchars < 0) {
htmlParseErr(ctxt, XML_ERR_INVALID_ENCODING, htmlParseErr(ctxt, XML_ERR_INVALID_ENCODING,
"htmlCheckEncoding: encoder error\n", "htmlCheckEncoding: encoder error\n",
NULL, NULL); NULL, NULL);
} }
ctxt->input->base = ctxt->input->base =
ctxt->input->cur = ctxt->input->buf->buffer->content; ctxt->input->cur = xmlBufContent(ctxt->input->buf->buffer);
ctxt->input->end = ctxt->input->end = xmlBufEnd(ctxt->input->buf->buffer);
&ctxt->input->base[ctxt->input->buf->buffer->use];
} }
} }
} }
@@ -4906,9 +4906,9 @@ htmlCreateMemoryParserCtxt(const char *buffer, int size) {
input->filename = NULL; input->filename = NULL;
input->buf = buf; input->buf = buf;
input->base = input->buf->buffer->content; input->cur =
input->cur = input->buf->buffer->content; input->base = xmlBufContent(input->buf->buffer);
input->end = &input->buf->buffer->content[input->buf->buffer->use]; input->end = xmlBufEnd(input->buf->buffer);
inputPush(ctxt, input); inputPush(ctxt, input);
return(ctxt); return(ctxt);
@@ -5025,8 +5025,8 @@ htmlParseLookupSequence(htmlParserCtxtPtr ctxt, xmlChar first,
buf = in->base; buf = in->base;
len = in->length; len = in->length;
} else { } else {
buf = in->buf->buffer->content; buf = xmlBufContent(in->buf->buffer);
len = in->buf->buffer->use; len = xmlBufUse(in->buf->buffer);
} }
/* take into account the sequence length */ /* take into account the sequence length */
@@ -5152,8 +5152,8 @@ htmlParseLookupChars(htmlParserCtxtPtr ctxt, const xmlChar * stop,
buf = in->base; buf = in->base;
len = in->length; len = in->length;
} else { } else {
buf = in->buf->buffer->content; buf = xmlBufContent(in->buf->buffer);
len = in->buf->buffer->use; len = xmlBufUse(in->buf->buffer);
} }
for (; base < len; base++) { for (; base < len; base++) {
@@ -5264,7 +5264,7 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
if (in->buf == NULL) if (in->buf == NULL)
avail = in->length - (in->cur - in->base); avail = in->length - (in->cur - in->base);
else else
avail = in->buf->buffer->use - (in->cur - in->base); avail = xmlBufUse(in->buf->buffer) - (in->cur - in->base);
if ((avail == 0) && (terminate)) { if ((avail == 0) && (terminate)) {
htmlAutoCloseOnEnd(ctxt); htmlAutoCloseOnEnd(ctxt);
if ((ctxt->nameNr == 0) && (ctxt->instate != XML_PARSER_EOF)) { if ((ctxt->nameNr == 0) && (ctxt->instate != XML_PARSER_EOF)) {
@@ -5300,7 +5300,7 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
if (in->buf == NULL) if (in->buf == NULL)
avail = in->length - (in->cur - in->base); avail = in->length - (in->cur - in->base);
else else
avail = in->buf->buffer->use - (in->cur - in->base); avail = xmlBufUse(in->buf->buffer) - (in->cur - in->base);
} }
if ((ctxt->sax) && (ctxt->sax->setDocumentLocator)) if ((ctxt->sax) && (ctxt->sax->setDocumentLocator))
ctxt->sax->setDocumentLocator(ctxt->userData, ctxt->sax->setDocumentLocator(ctxt->userData,
@@ -5342,7 +5342,7 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
if (in->buf == NULL) if (in->buf == NULL)
avail = in->length - (in->cur - in->base); avail = in->length - (in->cur - in->base);
else else
avail = in->buf->buffer->use - (in->cur - in->base); avail = xmlBufUse(in->buf->buffer) - (in->cur - in->base);
/* /*
* no chars in buffer * no chars in buffer
*/ */
@@ -5415,7 +5415,7 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
if (in->buf == NULL) if (in->buf == NULL)
avail = in->length - (in->cur - in->base); avail = in->length - (in->cur - in->base);
else else
avail = in->buf->buffer->use - (in->cur - in->base); avail = xmlBufUse(in->buf->buffer) - (in->cur - in->base);
if (avail < 2) if (avail < 2)
goto done; goto done;
cur = in->cur[0]; cur = in->cur[0];
@@ -5456,7 +5456,7 @@ htmlParseTryOrFinish(htmlParserCtxtPtr ctxt, int terminate) {
if (in->buf == NULL) if (in->buf == NULL)
avail = in->length - (in->cur - in->base); avail = in->length - (in->cur - in->base);
else else
avail = in->buf->buffer->use - (in->cur - in->base); avail = xmlBufUse(in->buf->buffer) - (in->cur - in->base);
if (avail < 1) if (avail < 1)
goto done; goto done;
cur = in->cur[0]; cur = in->cur[0];
@@ -5979,7 +5979,7 @@ htmlParseChunk(htmlParserCtxtPtr ctxt, const char *chunk, int size,
} }
if ((size > 0) && (chunk != NULL) && (ctxt->input != NULL) && if ((size > 0) && (chunk != NULL) && (ctxt->input != NULL) &&
(ctxt->input->buf != NULL) && (ctxt->instate != XML_PARSER_EOF)) { (ctxt->input->buf != NULL) && (ctxt->instate != XML_PARSER_EOF)) {
int base = ctxt->input->base - ctxt->input->buf->buffer->content; int base = ctxt->input->base - xmlBufContent(ctxt->input->buf->buffer);
int cur = ctxt->input->cur - ctxt->input->base; int cur = ctxt->input->cur - ctxt->input->base;
int res; int res;
@@ -5989,10 +5989,9 @@ htmlParseChunk(htmlParserCtxtPtr ctxt, const char *chunk, int size,
ctxt->disableSAX = 1; ctxt->disableSAX = 1;
return (XML_PARSER_EOF); return (XML_PARSER_EOF);
} }
ctxt->input->base = ctxt->input->buf->buffer->content + base; ctxt->input->base = xmlBufContent(ctxt->input->buf->buffer) + base;
ctxt->input->cur = ctxt->input->base + cur; ctxt->input->cur = ctxt->input->base + cur;
ctxt->input->end = ctxt->input->end = xmlBufEnd(ctxt->input->buf->buffer);
&ctxt->input->buf->buffer->content[ctxt->input->buf->buffer->use];
#ifdef DEBUG_PUSH #ifdef DEBUG_PUSH
xmlGenericError(xmlGenericErrorContext, "HPP: pushed %d\n", size); xmlGenericError(xmlGenericErrorContext, "HPP: pushed %d\n", size);
#endif #endif
@@ -6008,7 +6007,7 @@ htmlParseChunk(htmlParserCtxtPtr ctxt, const char *chunk, int size,
(in->raw != NULL)) { (in->raw != NULL)) {
int nbchars; int nbchars;
nbchars = xmlCharEncInFunc(in->encoder, in->buffer, in->raw); nbchars = xmlCharEncInput(in);
if (nbchars < 0) { if (nbchars < 0) {
htmlParseErr(ctxt, XML_ERR_INVALID_ENCODING, htmlParseErr(ctxt, XML_ERR_INVALID_ENCODING,
"encoder error\n", NULL, NULL); "encoder error\n", NULL, NULL);
@@ -6107,24 +6106,22 @@ htmlCreatePushParserCtxt(htmlSAXHandlerPtr sax, void *user_data,
inputStream->filename = (char *) inputStream->filename = (char *)
xmlCanonicPath((const xmlChar *) filename); xmlCanonicPath((const xmlChar *) filename);
inputStream->buf = buf; inputStream->buf = buf;
inputStream->base = inputStream->buf->buffer->content; inputStream->cur =
inputStream->cur = inputStream->buf->buffer->content; inputStream->base = xmlBufContent(buf->buffer);
inputStream->end = inputStream->end = xmlBufEnd(buf->buffer);
&inputStream->buf->buffer->content[inputStream->buf->buffer->use];
inputPush(ctxt, inputStream); inputPush(ctxt, inputStream);
if ((size > 0) && (chunk != NULL) && (ctxt->input != NULL) && if ((size > 0) && (chunk != NULL) && (ctxt->input != NULL) &&
(ctxt->input->buf != NULL)) { (ctxt->input->buf != NULL)) {
int base = ctxt->input->base - ctxt->input->buf->buffer->content; int base = ctxt->input->base - xmlBufContent(ctxt->input->buf->buffer);
int cur = ctxt->input->cur - ctxt->input->base; int cur = ctxt->input->cur - ctxt->input->base;
xmlParserInputBufferPush(ctxt->input->buf, size, chunk); xmlParserInputBufferPush(ctxt->input->buf, size, chunk);
ctxt->input->base = ctxt->input->buf->buffer->content + base; ctxt->input->base = xmlBufContent(ctxt->input->buf->buffer) + base;
ctxt->input->cur = ctxt->input->base + cur; ctxt->input->cur = ctxt->input->base + cur;
ctxt->input->end = ctxt->input->end = xmlBufEnd(ctxt->input->buf->buffer);
&ctxt->input->buf->buffer->content[ctxt->input->buf->buffer->use];
#ifdef DEBUG_PUSH #ifdef DEBUG_PUSH
xmlGenericError(xmlGenericErrorContext, "HPP: pushed %d\n", size); xmlGenericError(xmlGenericErrorContext, "HPP: pushed %d\n", size);
#endif #endif