diff --git a/ChangeLog b/ChangeLog index 23387202..e7976b7f 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,3 +1,9 @@ +Fri Jul 16 11:01:40 CEST 2004 Daniel Veillard + + * catalog.c test/catalogs/white* result/catalogs/white*: + applied patches from Peter Breitenlohner to fix handling + of white space normalization in public ids and add tests + Tue Jul 13 17:24:13 CEST 2004 Daniel Veillard * xmlmemory.c: applied a small fix from Steve Hay diff --git a/catalog.c b/catalog.c index 1ef78e7e..5dcc928b 100644 --- a/catalog.c +++ b/catalog.c @@ -75,6 +75,7 @@ void* __stdcall GetModuleHandleA(const char*); unsigned long __stdcall GetModuleFileNameA(void*, char*, unsigned long); #endif +static xmlChar *xmlCatalogNormalizePublic(const xmlChar *pubID); static int xmlExpandCatalog(xmlCatalogPtr catal, const char *filename); /************************************************************************ @@ -254,6 +255,7 @@ static xmlCatalogEntryPtr xmlNewCatalogEntry(xmlCatalogEntryType type, const xmlChar *name, const xmlChar *value, const xmlChar *URL, xmlCatalogPrefer prefer) { xmlCatalogEntryPtr ret; + xmlChar *normid = NULL; ret = (xmlCatalogEntryPtr) xmlMalloc(sizeof(xmlCatalogEntry)); if (ret == NULL) { @@ -264,10 +266,17 @@ xmlNewCatalogEntry(xmlCatalogEntryType type, const xmlChar *name, ret->parent = NULL; ret->children = NULL; ret->type = type; + if (type == XML_CATA_PUBLIC || type == XML_CATA_DELEGATE_PUBLIC) { + normid = xmlCatalogNormalizePublic(name); + if (normid != NULL) + name = (*normid != 0 ? normid : NULL); + } if (name != NULL) ret->name = xmlStrdup(name); else ret->name = NULL; + if (normid != NULL) + xmlFree(normid); if (value != NULL) ret->value = xmlStrdup(value); else @@ -943,6 +952,61 @@ xmlLoadFileContent(const char *filename) return(content); } +/** + * xmlCatalogNormalizePublic: + * @pubID: the public ID string + * + * Normalizes the Public Identifier + * + * Implements 6.2. Public Identifier Normalization + * from http://www.oasis-open.org/committees/entity/spec-2001-08-06.html + * + * Returns the new string or NULL, the string must be deallocated + * by the caller. + */ +static xmlChar * +xmlCatalogNormalizePublic(const xmlChar *pubID) +{ + int ok = 1; + int white; + const xmlChar *p; + xmlChar *ret; + xmlChar *q; + + if (pubID == NULL) + return(NULL); + + white = 1; + for (p = pubID;*p != 0 && ok;p++) { + if (!xmlIsBlank_ch(*p)) + white = 0; + else if (*p == 0x20 && !white) + white = 1; + else + ok = 0; + } + if (ok && !white) /* is normalized */ + return(NULL); + + ret = xmlStrdup(pubID); + q = ret; + white = 0; + for (p = pubID;*p != 0;p++) { + if (xmlIsBlank_ch(*p)) { + if (q != ret) + white = 1; + } else { + if (white) { + *(q++) = 0x20; + white = 0; + } + *(q++) = *p; + } + } + *q = 0; + return(ret); +} + /************************************************************************ * * * The XML Catalog parser * @@ -1858,12 +1922,17 @@ xmlCatalogListXMLResolve(xmlCatalogEntryPtr catal, const xmlChar *pubID, const xmlChar *sysID) { xmlChar *ret = NULL; xmlChar *urnID = NULL; + xmlChar *normid; if (catal == NULL) return(NULL); if ((pubID == NULL) && (sysID == NULL)) return(NULL); + normid = xmlCatalogNormalizePublic(pubID); + if (normid != NULL) + pubID = (*normid != 0 ? normid : NULL); + if (!xmlStrncmp(pubID, BAD_CAST XML_URN_PUBID, sizeof(XML_URN_PUBID) - 1)) { urnID = xmlCatalogUnWrapURN(pubID); if (xmlDebugCatalogs) { @@ -1877,6 +1946,8 @@ xmlCatalogListXMLResolve(xmlCatalogEntryPtr catal, const xmlChar *pubID, ret = xmlCatalogListXMLResolve(catal, urnID, sysID); if (urnID != NULL) xmlFree(urnID); + if (normid != NULL) + xmlFree(normid); return(ret); } if (!xmlStrncmp(sysID, BAD_CAST XML_URN_PUBID, sizeof(XML_URN_PUBID) - 1)) { @@ -1898,6 +1969,8 @@ xmlCatalogListXMLResolve(xmlCatalogEntryPtr catal, const xmlChar *pubID, } if (urnID != NULL) xmlFree(urnID); + if (normid != NULL) + xmlFree(normid); return(ret); } while (catal != NULL) { @@ -1907,12 +1980,17 @@ xmlCatalogListXMLResolve(xmlCatalogEntryPtr catal, const xmlChar *pubID, } if (catal->children != NULL) { ret = xmlCatalogXMLResolve(catal->children, pubID, sysID); - if (ret != NULL) + if (ret != NULL) { + if (normid != NULL) + xmlFree(normid); return(ret); + } } } catal = catal->next; } + if (normid != NULL) + xmlFree(normid); return(ret); } @@ -2143,8 +2221,6 @@ xmlGetSGMLCatalogEntryType(const xmlChar *name) { type = SGML_CATA_CATALOG; else if (xmlStrEqual(name, (const xmlChar *) "BASE")) type = SGML_CATA_BASE; - else if (xmlStrEqual(name, (const xmlChar *) "DELEGATE")) - type = SGML_CATA_DELEGATE; return(type); } @@ -2219,8 +2295,6 @@ xmlParseSGMLCatalog(xmlCatalogPtr catal, const xmlChar *value, type = SGML_CATA_CATALOG; else if (xmlStrEqual(name, (const xmlChar *) "BASE")) type = SGML_CATA_BASE; - else if (xmlStrEqual(name, (const xmlChar *) "DELEGATE")) - type = SGML_CATA_DELEGATE; else if (xmlStrEqual(name, (const xmlChar *) "OVERRIDE")) { xmlFree(name); cur = xmlParseSGMLCatalogName(cur, &name); @@ -2266,6 +2340,21 @@ xmlParseSGMLCatalog(xmlCatalogPtr catal, const xmlChar *value, /* error */ break; } + if (type != SGML_CATA_SYSTEM) { + xmlChar *normid; + + normid = xmlCatalogNormalizePublic(name); + if (normid != NULL) { + if (name != NULL) + xmlFree(name); + if (*normid != 0) + name = normid; + else { + xmlFree(normid); + name = NULL; + } + } + } if (!IS_BLANK_CH(*cur)) { /* error */ break; @@ -2371,15 +2460,28 @@ xmlParseSGMLCatalog(xmlCatalogPtr catal, const xmlChar *value, static const xmlChar * xmlCatalogGetSGMLPublic(xmlHashTablePtr catal, const xmlChar *pubID) { xmlCatalogEntryPtr entry; + xmlChar *normid; if (catal == NULL) return(NULL); + normid = xmlCatalogNormalizePublic(pubID); + if (normid != NULL) + pubID = (*normid != 0 ? normid : NULL); + entry = (xmlCatalogEntryPtr) xmlHashLookup(catal, pubID); - if (entry == NULL) + if (entry == NULL) { + if (normid != NULL) + xmlFree(normid); return(NULL); - if (entry->type == SGML_CATA_PUBLIC) + } + if (entry->type == SGML_CATA_PUBLIC) { + if (normid != NULL) + xmlFree(normid); return(entry->URL); + } + if (normid != NULL) + xmlFree(normid); return(NULL); } diff --git a/result/catalogs/whites b/result/catalogs/whites new file mode 100644 index 00000000..52b2afed --- /dev/null +++ b/result/catalogs/whites @@ -0,0 +1,5 @@ +> test/catalogs/calstblx.dtd +> test/catalogs/docbook.dtd +> test/catalogs/soextblx.dtd +> test/catalogs/dbgenent.mod +> \ No newline at end of file diff --git a/result/catalogs/whitex b/result/catalogs/whitex new file mode 100644 index 00000000..b02bcc60 --- /dev/null +++ b/result/catalogs/whitex @@ -0,0 +1,8 @@ +> /usr/share/xml/docbook/xml/4.1.2/dbpoolx.mod +> http://www.oasis-open.org/docbook/xml/4.1.2/dbcentx.mod +> http://www.oasis-open.org/docbook/xml/4.1.2/dbcentx.mod +> http://www.oasis-open.org/docbook/xml/4.1.2/dbcentx.mod +> http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd +> http://www.oasis-open.org/docbook/xml/4.1.2/docbookx.dtd +> file:///usr/share/xml/toto/toto.dtd +> \ No newline at end of file diff --git a/test/catalogs/whites.script b/test/catalogs/whites.script new file mode 100644 index 00000000..8f4dd139 --- /dev/null +++ b/test/catalogs/whites.script @@ -0,0 +1,4 @@ +public "-//OASIS//DTD DocBook XML CALS Table Model V4.1//EN " +public "-//Davenport//DTD DocBook V3.0//EN" +public " -//OASIS//DTD XML Exchange Table Model 19990315//EN" +public "-//Davenport//ENTITIES DocBook Additional General Entities V3.0//EN" diff --git a/test/catalogs/whites.sgml b/test/catalogs/whites.sgml new file mode 100644 index 00000000..50269b72 --- /dev/null +++ b/test/catalogs/whites.sgml @@ -0,0 +1,5 @@ +PUBLIC " -//OASIS//DTD DocBook XML CALS Table Model V4.1//EN" "calstblx.dtd" +PUBLIC "-//OASIS//DTD XML Exchange Table Model 19990315//EN " "soextblx.dtd" +PUBLIC "-//Davenport//DTD DocBook + V3.0//EN" "docbook.dtd" +PUBLIC "-//Davenport//ENTITIES DocBook Additional General Entities V3.0//EN" "dbgenent.mod" diff --git a/test/catalogs/whitex.script b/test/catalogs/whitex.script new file mode 100644 index 00000000..a49ca235 --- /dev/null +++ b/test/catalogs/whitex.script @@ -0,0 +1,7 @@ +resolve "toto " http://www.oasis-open.org/docbook/xml/4.1.2/dbpoolx.mod +public "-//OASIS//ENTITIES DocBook XML Character Entities V4.1.2//EN" +public " -//OASIS//ENTITIES DocBook XML Character Entities V4.1.2//EN" +public "-//OASIS//ENTITIES DocBook XML Character Entities V4.1.2//EN " +system urn:publicid:+-:OASIS:DTD+++DocBook+XML+V4.1.2:EN+ +public urn:publicid:+-:OASIS:DTD+DocBook+XML+++V4.1.2:EN+ +resolve " toto " toto diff --git a/test/catalogs/whitex.xml b/test/catalogs/whitex.xml new file mode 100644 index 00000000..d742887e --- /dev/null +++ b/test/catalogs/whitex.xml @@ -0,0 +1,15 @@ + + + + + + + + + + + + + +