From 5e873c4627c597cefd0472c85002ccab4f13afa7 Mon Sep 17 00:00:00 2001 From: Daniel Veillard Date: Wed, 12 Apr 2000 13:27:38 +0000 Subject: [PATCH] - speedup of IS_CHAR like macros, significant overall improvement - More interfaces for new I/O functions: xmlNewIOInputStream, xmlParserInputBufferCreateIO, xmlCreateIOParserCtxt - added I/O test to xmllint Daniel --- ChangeLog | 11 + doc/html/index.sgml | 1288 +++++++++++++++--------------- doc/upgrade.html | 18 +- include/libxml/parser.h | 15 + include/libxml/parserInternals.h | 16 +- include/libxml/xmlIO.h | 5 + parser.c | 118 ++- parser.h | 15 + parserInternals.h | 16 +- testSAX.c | 43 +- xmlIO.c | 29 + xmlIO.h | 5 + xmllint.c | 40 + 13 files changed, 936 insertions(+), 683 deletions(-) diff --git a/ChangeLog b/ChangeLog index 284b13db..370bb5a3 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,14 @@ +Wed Apr 12 15:47:22 CEST 2000 Daniel Veillard + + * xmlIO.[ch] parser.[ch]: More interfaces for new I/O functions + xmlNewIOInputStream, xmlParserInputBufferCreateIO, + xmlCreateIOParserCtxt + * parser.c parserInternals.h: speedup of IS_CHAR like macros, + significant overall improvement + * xmllint.c: added I/O test to xmllint + * testSAX.c: added a speed test + * doc/* : updated/regenerated + Sat Apr 8 14:54:54 CEST 2000 Daniel Veillard * xpath.c uri.h parserInternals.h: cosmetic changes from diff --git a/doc/html/index.sgml b/doc/html/index.sgml index 85363ee7..5776c723 100644 --- a/doc/html/index.sgml +++ b/doc/html/index.sgml @@ -1,644 +1,644 @@ - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/doc/upgrade.html b/doc/upgrade.html index 83c5190f..658de2a1 100644 --- a/doc/upgrade.html +++ b/doc/upgrade.html @@ -44,7 +44,17 @@ mail:

Use xmlDocGetRootElement(doc) to get the root element of a document. Alternatively if you are sure to not reference Dtds nor have PIs or comments before or after the root element s/->root/->children/g - will probably do it. + will probably do it. +

Note: libxml2 final version now export a version + number as the LIBXML_VERSION preprocessor token. In most case the changes + required for 1/ and 2/ can be dealt with using the following construct (if + you don't use root identifier for other purposes):

+ 
#if defined(LIBXML_VERSION) && LIBXML_VERSION >= 20000
+#define root children
+#define childs children
+#endif
+
+
  • The white space issue, this one is more complex, unless special case of validating parsing, the line breaks and spaces usually used for indenting and formatting the document content becomes significant. So they are @@ -69,11 +79,11 @@ mail:

  • The include path has changed to $prefix/libxml/ and the includes themselves uses this new prefix in includes instructions... If you are - using (as expected) the + using (as expected) the 
    xml-config --cflags

    output to generate you compile commands this will probably work out of the box

    -

    +

    • @@ -85,6 +95,6 @@ upgrade, it may cost a lot on the long term ...

    Daniel Veillard

    -

    $Id: upgrade.html,v 1.2 2000/03/06 07:41:49 veillard Exp $

    +

    $Id: upgrade.html,v 1.3 2000/04/03 19:48:13 veillard Exp $

    diff --git a/include/libxml/parser.h b/include/libxml/parser.h index d755b70b..aa7e2735 100644 --- a/include/libxml/parser.h +++ b/include/libxml/parser.h @@ -428,6 +428,21 @@ int xmlParseChunk (xmlParserCtxtPtr ctxt, int size, int terminate); +/** + * Special I/O mode + */ + +xmlParserCtxtPtr xmlCreateIOParserCtxt (xmlSAXHandlerPtr sax, + void *user_data, + xmlInputReadCallback ioread, + xmlInputCloseCallback ioclose, + void *ioctx, + xmlCharEncoding enc); + +xmlParserInputPtr xmlNewIOInputStream (xmlParserCtxtPtr ctxt, + xmlParserInputBufferPtr input, + xmlCharEncoding enc); + /** * Node infos */ diff --git a/include/libxml/parserInternals.h b/include/libxml/parserInternals.h index 0972de42..c3597348 100644 --- a/include/libxml/parserInternals.h +++ b/include/libxml/parserInternals.h @@ -53,6 +53,7 @@ extern "C" { (((c) >= 0x00C0) && ((c) <= 0x00D6)) || \ (((c) >= 0x00D8) && ((c) <= 0x00F6)) || \ (((c) >= 0x00F8) && ((c) <= 0x00FF)) || \ + (((c) >= 0x100) && ( /* accelerator */ \ (((c) >= 0x0100) && ((c) <= 0x0131)) || \ (((c) >= 0x0134) && ((c) <= 0x013E)) || \ (((c) >= 0x0141) && ((c) <= 0x0148)) || \ @@ -97,6 +98,7 @@ extern "C" { (((c) >= 0x06D0) && ((c) <= 0x06D3)) || \ ((c) == 0x06D5) || \ (((c) >= 0x06E5) && ((c) <= 0x06E6)) || \ + (((c) > 0x905) && ( /* accelerator */ \ (((c) >= 0x0905) && ((c) <= 0x0939)) || \ ((c) == 0x093D) || \ (((c) >= 0x0958) && ((c) <= 0x0961)) || \ @@ -187,6 +189,7 @@ extern "C" { (((c) >= 0x0EC0) && ((c) <= 0x0EC4)) || \ (((c) >= 0x0F40) && ((c) <= 0x0F47)) || \ (((c) >= 0x0F49) && ((c) <= 0x0F69)) || \ + (((c) > 0x10A0) && ( /* accelerator */ \ (((c) >= 0x10A0) && ((c) <= 0x10C5)) || \ (((c) >= 0x10D0) && ((c) <= 0x10F6)) || \ ((c) == 0x1100) || \ @@ -249,13 +252,14 @@ extern "C" { (((c) >= 0x3041) && ((c) <= 0x3094)) || \ (((c) >= 0x30A1) && ((c) <= 0x30FA)) || \ (((c) >= 0x3105) && ((c) <= 0x312C)) || \ - (((c) >= 0xAC00) && ((c) <= 0xD7A3))) + (((c) >= 0xAC00) && ((c) <= 0xD7A3))) /* accelerators */ )))))) /* * [88] Digit ::= ... long list see REC ... */ #define IS_DIGIT(c) \ ((((c) >= 0x0030) && ((c) <= 0x0039)) || \ + (((c) >= 0x660) && ( /* accelerator */ \ (((c) >= 0x0660) && ((c) <= 0x0669)) || \ (((c) >= 0x06F0) && ((c) <= 0x06F9)) || \ (((c) >= 0x0966) && ((c) <= 0x096F)) || \ @@ -269,13 +273,14 @@ extern "C" { (((c) >= 0x0D66) && ((c) <= 0x0D6F)) || \ (((c) >= 0x0E50) && ((c) <= 0x0E59)) || \ (((c) >= 0x0ED0) && ((c) <= 0x0ED9)) || \ - (((c) >= 0x0F20) && ((c) <= 0x0F29))) + (((c) >= 0x0F20) && ((c) <= 0x0F29))) /* accelerator */ )) /* * [87] CombiningChar ::= ... long list see REC ... */ #define IS_COMBINING(c) \ - ((((c) >= 0x0300) && ((c) <= 0x0345)) || \ + ((((c) >= 0x300) && ( /* accelerator */ \ + (((c) >= 0x0300) && ((c) <= 0x0345)) || \ (((c) >= 0x0360) && ((c) <= 0x0361)) || \ (((c) >= 0x0483) && ((c) <= 0x0486)) || \ (((c) >= 0x0591) && ((c) <= 0x05A1)) || \ @@ -291,6 +296,7 @@ extern "C" { (((c) >= 0x06E0) && ((c) <= 0x06E4)) || \ (((c) >= 0x06E7) && ((c) <= 0x06E8)) || \ (((c) >= 0x06EA) && ((c) <= 0x06ED)) || \ + (((c) > 0x0901) && ( /* accelerator */ \ (((c) >= 0x0901) && ((c) <= 0x0903)) || \ ((c) == 0x093C) || \ (((c) >= 0x093E) && ((c) <= 0x094C)) || \ @@ -306,6 +312,7 @@ extern "C" { (((c) >= 0x09CB) && ((c) <= 0x09CD)) || \ ((c) == 0x09D7) || \ (((c) >= 0x09E2) && ((c) <= 0x09E3)) || \ + (((c) > 0x0A02) && ( /* accelerator */ \ ((c) == 0x0A02) || \ ((c) == 0x0A3C) || \ ((c) == 0x0A3E) || \ @@ -345,6 +352,7 @@ extern "C" { (((c) >= 0x0D46) && ((c) <= 0x0D48)) || \ (((c) >= 0x0D4A) && ((c) <= 0x0D4D)) || \ ((c) == 0x0D57) || \ + (((c) > 0x0E31) && ( /* accelerator */ \ ((c) == 0x0E31) || \ (((c) >= 0x0E34) && ((c) <= 0x0E3A)) || \ (((c) >= 0x0E47) && ((c) <= 0x0E4E)) || \ @@ -369,7 +377,7 @@ extern "C" { ((c) == 0x20E1) || \ (((c) >= 0x302A) && ((c) <= 0x302F)) || \ ((c) == 0x3099) || \ - ((c) == 0x309A)) + ((c) == 0x309A)) /* accelerator */ )))))))) /* * [89] Extender ::= #x00B7 | #x02D0 | #x02D1 | #x0387 | #x0640 | diff --git a/include/libxml/xmlIO.h b/include/libxml/xmlIO.h index 0722e709..bc1c63b3 100644 --- a/include/libxml/xmlIO.h +++ b/include/libxml/xmlIO.h @@ -52,6 +52,11 @@ xmlParserInputBufferPtr xmlParserInputBufferPtr xmlParserInputBufferCreateFd (int fd, xmlCharEncoding enc); +xmlParserInputBufferPtr + xmlParserInputBufferCreateIO (xmlInputReadCallback ioread, + xmlInputCloseCallback ioclose, + void *ioctx, + xmlCharEncoding enc); int xmlParserInputBufferRead (xmlParserInputBufferPtr in, int len); int xmlParserInputBufferGrow (xmlParserInputBufferPtr in, diff --git a/parser.c b/parser.c index a10554e7..9b6ec392 100644 --- a/parser.c +++ b/parser.c @@ -921,11 +921,13 @@ xmlNewInputStream(xmlParserCtxtPtr ctxt) { input = (xmlParserInputPtr) xmlMalloc(sizeof(xmlParserInput)); if (input == NULL) { - ctxt->errNo = XML_ERR_NO_MEMORY; - if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL)) - ctxt->sax->error(ctxt->userData, - "malloc: couldn't allocate a new input stream\n"); - ctxt->errNo = XML_ERR_NO_MEMORY; + if (ctxt != NULL) { + ctxt->errNo = XML_ERR_NO_MEMORY; + if ((ctxt->sax != NULL) && (ctxt->sax->error != NULL)) + ctxt->sax->error(ctxt->userData, + "malloc: couldn't allocate a new input stream\n"); + ctxt->errNo = XML_ERR_NO_MEMORY; + } return(NULL); } memset(input, 0, sizeof(xmlParserInput)); @@ -935,6 +937,37 @@ xmlNewInputStream(xmlParserCtxtPtr ctxt) { return(input); } +/** + * xmlNewIOInputStream: + * @ctxt: an XML parser context + * @input: an I/O Input + * @enc: the charset encoding if known + * + * Create a new input stream structure encapsulating the @input into + * a stream suitable for the parser. + * + * Returns the new input stream or NULL + */ +xmlParserInputPtr +xmlNewIOInputStream(xmlParserCtxtPtr ctxt, xmlParserInputBufferPtr input, + xmlCharEncoding enc) { + xmlParserInputPtr inputStream; + + inputStream = xmlNewInputStream(ctxt); + if (inputStream == NULL) { + return(NULL); + } + inputStream->filename = NULL; + inputStream->buf = input; + inputStream->base = inputStream->buf->buffer->content; + inputStream->cur = inputStream->buf->buffer->content; + if (enc != XML_CHAR_ENCODING_NONE) { + xmlSwitchEncoding(ctxt, enc); + } + + return(inputStream); +} + /** * xmlNewEntityInputStream: * @ctxt: an XML parser context @@ -3403,16 +3436,18 @@ xmlParseName(xmlParserCtxtPtr ctxt) { GROW; c = CUR_CHAR(l); - if (!IS_LETTER(c) && (c != '_') && - (c != ':')) { + if ((c == ' ') || (c == '>') || (c == '/') || /* accelerators */ + (!IS_LETTER(c) && (c != '_') && + (c != ':'))) { return(NULL); } - while ((IS_LETTER(c)) || (IS_DIGIT(c)) || - (c == '.') || (c == '-') || - (c == '_') || (c == ':') || - (IS_COMBINING(c)) || - (IS_EXTENDER(c))) { + while ((c != ' ') && (c != '>') && (c != '/') && /* accelerators */ + ((IS_LETTER(c)) || (IS_DIGIT(c)) || + (c == '.') || (c == '-') || + (c == '_') || (c == ':') || + (IS_COMBINING(c)) || + (IS_EXTENDER(c)))) { COPY_BUF(l,buf,len,c); NEXTL(l); c = CUR_CHAR(l); @@ -4041,8 +4076,9 @@ xmlParseCharData(xmlParserCtxtPtr ctxt, int cdata) { SHRINK; cur = CUR_CHAR(l); - while ((IS_CHAR(cur)) && ((cur != '<') || (ctxt->token == '<')) && - ((cur != '&') || (ctxt->token == '&'))) { + while (((cur != '<') || (ctxt->token == '<')) && + ((cur != '&') || (ctxt->token == '&')) && + (IS_CHAR(cur))) { if ((cur == ']') && (NXT(1) == ']') && (NXT(2) == '>')) { if (cdata) break; @@ -9524,6 +9560,60 @@ xmlCreatePushParserCtxt(xmlSAXHandlerPtr sax, void *user_data, return(ctxt); } +/** + * xmlCreateIOParserCtxt: + * @sax: a SAX handler + * @user_data: The user data returned on SAX callbacks + * @ioread: an I/O read function + * @ioclose: an I/O close function + * @ioctx: an I/O handler + * @enc: the charset encoding if known + * + * Create a parser context for using the XML parser with an existing + * I/O stream + * + * Returns the new parser context or NULL + */ +xmlParserCtxtPtr +xmlCreateIOParserCtxt(xmlSAXHandlerPtr sax, void *user_data, + xmlInputReadCallback ioread, xmlInputCloseCallback ioclose, + void *ioctx, xmlCharEncoding enc) { + xmlParserCtxtPtr ctxt; + xmlParserInputPtr inputStream; + xmlParserInputBufferPtr buf; + + buf = xmlParserInputBufferCreateIO(ioread, ioclose, ioctx, enc); + if (buf == NULL) return(NULL); + + ctxt = xmlNewParserCtxt(); + if (ctxt == NULL) { + xmlFree(buf); + return(NULL); + } + if (sax != NULL) { + if (ctxt->sax != &xmlDefaultSAXHandler) + xmlFree(ctxt->sax); + ctxt->sax = (xmlSAXHandlerPtr) xmlMalloc(sizeof(xmlSAXHandler)); + if (ctxt->sax == NULL) { + xmlFree(buf); + xmlFree(ctxt); + return(NULL); + } + memcpy(ctxt->sax, sax, sizeof(xmlSAXHandler)); + if (user_data != NULL) + ctxt->userData = user_data; + } + + inputStream = xmlNewIOInputStream(ctxt, buf, enc); + if (inputStream == NULL) { + xmlFreeParserCtxt(ctxt); + return(NULL); + } + inputPush(ctxt, inputStream); + + return(ctxt); +} + /** * xmlCreateDocParserCtxt: * @cur: a pointer to an array of xmlChar diff --git a/parser.h b/parser.h index d755b70b..aa7e2735 100644 --- a/parser.h +++ b/parser.h @@ -428,6 +428,21 @@ int xmlParseChunk (xmlParserCtxtPtr ctxt, int size, int terminate); +/** + * Special I/O mode + */ + +xmlParserCtxtPtr xmlCreateIOParserCtxt (xmlSAXHandlerPtr sax, + void *user_data, + xmlInputReadCallback ioread, + xmlInputCloseCallback ioclose, + void *ioctx, + xmlCharEncoding enc); + +xmlParserInputPtr xmlNewIOInputStream (xmlParserCtxtPtr ctxt, + xmlParserInputBufferPtr input, + xmlCharEncoding enc); + /** * Node infos */ diff --git a/parserInternals.h b/parserInternals.h index 0972de42..c3597348 100644 --- a/parserInternals.h +++ b/parserInternals.h @@ -53,6 +53,7 @@ extern "C" { (((c) >= 0x00C0) && ((c) <= 0x00D6)) || \ (((c) >= 0x00D8) && ((c) <= 0x00F6)) || \ (((c) >= 0x00F8) && ((c) <= 0x00FF)) || \ + (((c) >= 0x100) && ( /* accelerator */ \ (((c) >= 0x0100) && ((c) <= 0x0131)) || \ (((c) >= 0x0134) && ((c) <= 0x013E)) || \ (((c) >= 0x0141) && ((c) <= 0x0148)) || \ @@ -97,6 +98,7 @@ extern "C" { (((c) >= 0x06D0) && ((c) <= 0x06D3)) || \ ((c) == 0x06D5) || \ (((c) >= 0x06E5) && ((c) <= 0x06E6)) || \ + (((c) > 0x905) && ( /* accelerator */ \ (((c) >= 0x0905) && ((c) <= 0x0939)) || \ ((c) == 0x093D) || \ (((c) >= 0x0958) && ((c) <= 0x0961)) || \ @@ -187,6 +189,7 @@ extern "C" { (((c) >= 0x0EC0) && ((c) <= 0x0EC4)) || \ (((c) >= 0x0F40) && ((c) <= 0x0F47)) || \ (((c) >= 0x0F49) && ((c) <= 0x0F69)) || \ + (((c) > 0x10A0) && ( /* accelerator */ \ (((c) >= 0x10A0) && ((c) <= 0x10C5)) || \ (((c) >= 0x10D0) && ((c) <= 0x10F6)) || \ ((c) == 0x1100) || \ @@ -249,13 +252,14 @@ extern "C" { (((c) >= 0x3041) && ((c) <= 0x3094)) || \ (((c) >= 0x30A1) && ((c) <= 0x30FA)) || \ (((c) >= 0x3105) && ((c) <= 0x312C)) || \ - (((c) >= 0xAC00) && ((c) <= 0xD7A3))) + (((c) >= 0xAC00) && ((c) <= 0xD7A3))) /* accelerators */ )))))) /* * [88] Digit ::= ... long list see REC ... */ #define IS_DIGIT(c) \ ((((c) >= 0x0030) && ((c) <= 0x0039)) || \ + (((c) >= 0x660) && ( /* accelerator */ \ (((c) >= 0x0660) && ((c) <= 0x0669)) || \ (((c) >= 0x06F0) && ((c) <= 0x06F9)) || \ (((c) >= 0x0966) && ((c) <= 0x096F)) || \ @@ -269,13 +273,14 @@ extern "C" { (((c) >= 0x0D66) && ((c) <= 0x0D6F)) || \ (((c) >= 0x0E50) && ((c) <= 0x0E59)) || \ (((c) >= 0x0ED0) && ((c) <= 0x0ED9)) || \ - (((c) >= 0x0F20) && ((c) <= 0x0F29))) + (((c) >= 0x0F20) && ((c) <= 0x0F29))) /* accelerator */ )) /* * [87] CombiningChar ::= ... long list see REC ... */ #define IS_COMBINING(c) \ - ((((c) >= 0x0300) && ((c) <= 0x0345)) || \ + ((((c) >= 0x300) && ( /* accelerator */ \ + (((c) >= 0x0300) && ((c) <= 0x0345)) || \ (((c) >= 0x0360) && ((c) <= 0x0361)) || \ (((c) >= 0x0483) && ((c) <= 0x0486)) || \ (((c) >= 0x0591) && ((c) <= 0x05A1)) || \ @@ -291,6 +296,7 @@ extern "C" { (((c) >= 0x06E0) && ((c) <= 0x06E4)) || \ (((c) >= 0x06E7) && ((c) <= 0x06E8)) || \ (((c) >= 0x06EA) && ((c) <= 0x06ED)) || \ + (((c) > 0x0901) && ( /* accelerator */ \ (((c) >= 0x0901) && ((c) <= 0x0903)) || \ ((c) == 0x093C) || \ (((c) >= 0x093E) && ((c) <= 0x094C)) || \ @@ -306,6 +312,7 @@ extern "C" { (((c) >= 0x09CB) && ((c) <= 0x09CD)) || \ ((c) == 0x09D7) || \ (((c) >= 0x09E2) && ((c) <= 0x09E3)) || \ + (((c) > 0x0A02) && ( /* accelerator */ \ ((c) == 0x0A02) || \ ((c) == 0x0A3C) || \ ((c) == 0x0A3E) || \ @@ -345,6 +352,7 @@ extern "C" { (((c) >= 0x0D46) && ((c) <= 0x0D48)) || \ (((c) >= 0x0D4A) && ((c) <= 0x0D4D)) || \ ((c) == 0x0D57) || \ + (((c) > 0x0E31) && ( /* accelerator */ \ ((c) == 0x0E31) || \ (((c) >= 0x0E34) && ((c) <= 0x0E3A)) || \ (((c) >= 0x0E47) && ((c) <= 0x0E4E)) || \ @@ -369,7 +377,7 @@ extern "C" { ((c) == 0x20E1) || \ (((c) >= 0x302A) && ((c) <= 0x302F)) || \ ((c) == 0x3099) || \ - ((c) == 0x309A)) + ((c) == 0x309A)) /* accelerator */ )))))))) /* * [89] Extender ::= #x00B7 | #x02D0 | #x02D1 | #x0387 | #x0640 | diff --git a/testSAX.c b/testSAX.c index d5079afb..0702f016 100644 --- a/testSAX.c +++ b/testSAX.c @@ -46,6 +46,7 @@ static int debug = 0; static int copy = 0; static int recovery = 0; static int push = 0; +static int speed = 0; xmlSAXHandler emptySAXHandlerStruct = { NULL, /* internalSubset */ @@ -635,20 +636,33 @@ void parseAndPrintFile(char *filename) { fclose(f); } } else { - /* - * Empty callbacks for checking - */ - res = xmlSAXUserParseFile(emptySAXHandler, NULL, filename); - if (res != 0) { - fprintf(stdout, "xmlSAXUserParseFile returned error %d\n", res); - } + if (!speed) { + /* + * Empty callbacks for checking + */ + res = xmlSAXUserParseFile(emptySAXHandler, NULL, filename); + if (res != 0) { + fprintf(stdout, "xmlSAXUserParseFile returned error %d\n", res); + } - /* - * Debug callback - */ - res = xmlSAXUserParseFile(debugSAXHandler, NULL, filename); - if (res != 0) { - fprintf(stdout, "xmlSAXUserParseFile returned error %d\n", res); + /* + * Debug callback + */ + res = xmlSAXUserParseFile(debugSAXHandler, NULL, filename); + if (res != 0) { + fprintf(stdout, "xmlSAXUserParseFile returned error %d\n", res); + } + } else { + /* + * test 100x the SAX parse + */ + int i; + + for (i = 0; i<100;i++) + res = xmlSAXUserParseFile(emptySAXHandler, NULL, filename); + if (res != 0) { + fprintf(stdout, "xmlSAXUserParseFile returned error %d\n", res); + } } } } @@ -669,6 +683,9 @@ int main(int argc, char **argv) { else if ((!strcmp(argv[i], "-push")) || (!strcmp(argv[i], "--push"))) push++; + else if ((!strcmp(argv[i], "-speed")) || + (!strcmp(argv[i], "--speed"))) + speed++; } for (i = 1; i < argc ; i++) { if (argv[i][0] != '-') { diff --git a/xmlIO.c b/xmlIO.c index ed73ce7c..65f5632b 100644 --- a/xmlIO.c +++ b/xmlIO.c @@ -635,6 +635,35 @@ xmlParserInputBufferCreateFd(int fd, xmlCharEncoding enc) { return(ret); } +/** + * xmlParserInputBufferCreateIO: + * @ioread: an I/O read function + * @ioclose: an I/O close function + * @ioctx: an I/O handler + * @enc: the charset encoding if known + * + * Create a buffered parser input for the progressive parsing for the input + * from a file descriptor + * + * Returns the new parser input or NULL + */ +xmlParserInputBufferPtr +xmlParserInputBufferCreateIO(xmlInputReadCallback ioread, + xmlInputCloseCallback ioclose, void *ioctx, xmlCharEncoding enc) { + xmlParserInputBufferPtr ret; + + if (ioread == NULL) return(NULL); + + ret = xmlAllocParserInputBuffer(enc); + if (ret != NULL) { + ret->context = (void *) ioctx; + ret->readcallback = ioread; + ret->closecallback = ioclose; + } + + return(ret); +} + /** * xmlParserInputBufferPush: * @in: a buffered parser input diff --git a/xmlIO.h b/xmlIO.h index 0722e709..bc1c63b3 100644 --- a/xmlIO.h +++ b/xmlIO.h @@ -52,6 +52,11 @@ xmlParserInputBufferPtr xmlParserInputBufferPtr xmlParserInputBufferCreateFd (int fd, xmlCharEncoding enc); +xmlParserInputBufferPtr + xmlParserInputBufferCreateIO (xmlInputReadCallback ioread, + xmlInputCloseCallback ioclose, + void *ioctx, + xmlCharEncoding enc); int xmlParserInputBufferRead (xmlParserInputBufferPtr in, int len); int xmlParserInputBufferGrow (xmlParserInputBufferPtr in, diff --git a/xmllint.c b/xmllint.c index aa58625c..24c17f71 100644 --- a/xmllint.c +++ b/xmllint.c @@ -67,6 +67,7 @@ static int html = 0; static int htmlout = 0; static int push = 0; static int noblanks = 0; +static int testIO = 0; extern int xmlDoValidityCheckingDefaultValue; extern int xmlGetWarningsDefaultValue; @@ -336,6 +337,19 @@ xmlShellReadline(char *prompt) { #endif } +/************************************************************************ + * * + * I/O Interfaces * + * * + ************************************************************************/ + +int myRead(FILE *f, char * buffer, int len) { + return(fread(buffer, 1, len, f)); +} +void myClose(FILE *f) { + fclose(f); +} + /************************************************************************ * * * Test processing * @@ -375,6 +389,28 @@ void parseAndPrintFile(char *filename) { xmlFreeParserCtxt(ctxt); } } + } else if (testIO) { + int ret; + FILE *f; + + f = fopen(filename, "r"); + if (f != NULL) { + xmlParserCtxtPtr ctxt; + + ctxt = xmlCreateIOParserCtxt(NULL, NULL, + (xmlInputReadCallback) myRead, + (xmlInputCloseCallback) myClose, + f, XML_CHAR_ENCODING_NONE); + xmlParseDocument(ctxt); + + ret = ctxt->wellFormed; + doc = ctxt->myDoc; + xmlFreeParserCtxt(ctxt); + if (!ret) { + xmlFreeDoc(doc); + doc = NULL; + } + } } else if (recovery) { doc = xmlRecoverFile(filename); } else if (htmlout) { @@ -545,6 +581,9 @@ int main(int argc, char **argv) { else if ((!strcmp(argv[i], "-push")) || (!strcmp(argv[i], "--push"))) push++; + else if ((!strcmp(argv[i], "-testIO")) || + (!strcmp(argv[i], "--testIO"))) + testIO++; else if ((!strcmp(argv[i], "-compress")) || (!strcmp(argv[i], "--compress"))) { compress++; @@ -612,6 +651,7 @@ int main(int argc, char **argv) { printf("\t--push : use the push mode of the parser\n"); printf("\t--nowarning : do not emit warnings from parser/validator\n"); printf("\t--noblanks : drop (ignorable?) blanks spaces\n"); + printf("\t--testIO : test user I/O support\n"); } xmlCleanupParser(); xmlMemoryDump();