X-Git-Url: https://www.fi.muni.cz/~kas/git//home/kas/public_html/git/?a=blobdiff_plain;f=backend%2Fpdf%2Fev-poppler.cc;h=917529e100e92daf80e47938f10fe8e5c5c1ac30;hb=5f554d8e362a35e05840068e1b3ea935be5d3c47;hp=0333d715564810d5209bc9856e0c2386fc8b6733;hpb=3d3328b2597e9b85afcb552ffc6abdeca1e3e3a7;p=evince.git diff --git a/backend/pdf/ev-poppler.cc b/backend/pdf/ev-poppler.cc index 0333d715..917529e1 100644 --- a/backend/pdf/ev-poppler.cc +++ b/backend/pdf/ev-poppler.cc @@ -1,5 +1,7 @@ /* -*- Mode: C++; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8; c-indent-level: 8 -*- */ -/* pdfdocument.h: Implementation of EvDocument for PDF +/* this file is part of evince, a gnome document viewer + * + * Copyright (C) 2009, Juanjo Marín * Copyright (C) 2004, Red Hat, Inc. * * This program is free software; you can redistribute it and/or modify @@ -63,6 +65,13 @@ #define HAVE_CAIRO_PRINT #endif +/* fields from the XMP Rights Management Schema, XMP Specification Sept 2005, pag. 45 */ +#define LICENSE_MARKED "/x:xmpmeta/rdf:RDF/rdf:Description/xmpRights:Marked" +#define LICENSE_TEXT "/x:xmpmeta/rdf:RDF/rdf:Description/dc:rights/rdf:Alt/rdf:li[lang('%s')]" +#define LICENSE_WEB_STATEMENT "/x:xmpmeta/rdf:RDF/rdf:Description/xmpRights:WebStatement" +/* license field from Creative Commons schema, http://creativecommons.org/ns */ +#define LICENSE_URI "/x:xmpmeta/rdf:RDF/rdf:Description/cc:license/@rdf:resource" + typedef struct { PdfDocument *document; char *text; @@ -92,12 +101,12 @@ typedef struct { struct _PdfDocumentClass { - GObjectClass parent_class; + EvDocumentClass parent_class; }; struct _PdfDocument { - GObject parent_instance; + EvDocument parent_instance; PopplerDocument *document; gchar *password; @@ -113,7 +122,6 @@ struct _PdfDocument GList *layers; }; -static void pdf_document_document_iface_init (EvDocumentIface *iface); static void pdf_document_security_iface_init (EvDocumentSecurityIface *iface); static void pdf_document_document_thumbnails_iface_init (EvDocumentThumbnailsIface *iface); static void pdf_document_document_links_iface_init (EvDocumentLinksIface *iface); @@ -233,14 +241,6 @@ pdf_document_dispose (GObject *object) G_OBJECT_CLASS (pdf_document_parent_class)->dispose (object); } -static void -pdf_document_class_init (PdfDocumentClass *klass) -{ - GObjectClass *g_object_class = G_OBJECT_CLASS (klass); - - g_object_class->dispose = pdf_document_dispose; -} - static void pdf_document_init (PdfDocument *pdf_document) { @@ -448,51 +448,18 @@ pdf_document_render (EvDocument *document, width, height, rc); } -/* EvDocumentSecurity */ - -static gboolean -pdf_document_has_document_security (EvDocumentSecurity *document_security) -{ - /* FIXME: do we really need to have this? */ - return FALSE; -} - -static void -pdf_document_set_password (EvDocumentSecurity *document_security, - const char *password) -{ - PdfDocument *document = PDF_DOCUMENT (document_security); - - if (document->password) - g_free (document->password); - - document->password = g_strdup (password); -} - - /* reference: http://www.pdfa.org/lib/exe/fetch.php?id=pdfa%3Aen%3Atechdoc&cache=cache&media=pdfa:techdoc:tn0001_pdfa-1_and_namespaces_2008-03-18.pdf */ static char * -pdf_document_get_format_from_metadata (const char *metadata) +pdf_document_get_format_from_metadata (xmlDocPtr doc, + xmlXPathContextPtr xpathCtx) { - xmlDocPtr doc; - xmlXPathContextPtr xpathCtx; xmlXPathObjectPtr xpathObj; xmlChar *part = NULL; xmlChar *conf = NULL; char *result = NULL; int i; - doc = xmlParseMemory (metadata, strlen (metadata)); - if (doc == NULL) - return NULL; /* invalid xml metadata */ - - xpathCtx = xmlXPathNewContext (doc); - if (xpathCtx == NULL) { - xmlFreeDoc (doc); - return NULL; /* invalid xpath context */ - } - /* add pdf/a namespaces */ xmlXPathRegisterNs (xpathCtx, BAD_CAST "x", BAD_CAST "adobe:ns:meta/"); xmlXPathRegisterNs (xpathCtx, BAD_CAST "rdf", BAD_CAST "http://www.w3.org/1999/02/22-rdf-syntax-ns#"); @@ -550,12 +517,159 @@ pdf_document_get_format_from_metadata (const char *metadata) /* Cleanup */ xmlFree (part); xmlFree (conf); - xmlXPathFreeContext (xpathCtx); - xmlFreeDoc (doc); return result; } +static EvDocumentLicense * +pdf_document_get_license_from_metadata (xmlDocPtr doc, + xmlXPathContextPtr xpathCtx) +{ + xmlXPathObjectPtr xpathObj; + xmlChar *marked = NULL; + const char *language_string; + char *aux; + gchar **tags; + gchar *tag, *tag_aux; + int i, j; + EvDocumentLicense *license; + + /* register namespaces */ + xmlXPathRegisterNs (xpathCtx, BAD_CAST "x", BAD_CAST "adobe:ns:meta/"); + xmlXPathRegisterNs (xpathCtx, BAD_CAST "rdf", BAD_CAST "http://www.w3.org/1999/02/22-rdf-syntax-ns#"); + xmlXPathRegisterNs (xpathCtx, BAD_CAST "dc", BAD_CAST "http://purl.org/dc/elements/1.1/"); + /* XMP Rights Management Schema */ + xmlXPathRegisterNs (xpathCtx, BAD_CAST "xmpRights", BAD_CAST "http://ns.adobe.com/xap/1.0/rights/"); + /* Creative Commons Schema */ + xmlXPathRegisterNs (xpathCtx, BAD_CAST "cc", BAD_CAST "http://creativecommons.org/ns#"); + + /* checking if the document has been marked as defined on the XMP Rights + * Management Schema */ + xpathObj = xmlXPathEvalExpression (BAD_CAST LICENSE_MARKED, xpathCtx); + if (xpathObj != NULL) { + if (xpathObj->nodesetval != NULL && + xpathObj->nodesetval->nodeNr != 0) + marked = xmlNodeGetContent (xpathObj->nodesetval->nodeTab[0]); + xmlXPathFreeObject (xpathObj); + } + + /* a) Not marked => No XMP Rights information */ + if (!marked) { + xmlFree (marked); + return NULL; + } + + license = ev_document_license_new (); + + /* b) Marked False => Public Domain, no copyrighted material and no + * license needed */ + if (g_strrstr ((char *) marked, "False") != NULL) { + license->text = g_strdup (_("This work is in the Public Domain")); + /* c) Marked True => Copyrighted material */ + } else { + /* Checking usage terms as defined by the XMP Rights Management + * Schema. This field is recomended to be checked by Creative + * Commons */ + /* 1) checking for a suitable localized string */ + language_string = pango_language_to_string (gtk_get_default_language ()); + tags = g_strsplit (language_string, "-", -1); + i = g_strv_length (tags); + while (i-- && !license->text) { + tag = g_strdup (tags[0]); + for (j = 1; j <= i; j++) { + tag_aux = g_strdup_printf ("%s-%s", tag, tags[j]); + g_free (tag); + tag = tag_aux; + } + aux = g_strdup_printf (LICENSE_TEXT, tag); + xpathObj = xmlXPathEvalExpression (BAD_CAST aux, xpathCtx); + if (xpathObj != NULL) { + if (xpathObj->nodesetval != NULL && + xpathObj->nodesetval->nodeNr != 0) + license->text = (gchar *)xmlNodeGetContent (xpathObj->nodesetval->nodeTab[0]); + xmlXPathFreeObject (xpathObj); + } + g_free (tag); + g_free (aux); + } + g_strfreev(tags); + + /* 2) if not, use the default string */ + if (!license->text) { + aux = g_strdup_printf (LICENSE_TEXT, "x-default"); + xpathObj = xmlXPathEvalExpression (BAD_CAST aux, xpathCtx); + if (xpathObj != NULL) { + if (xpathObj->nodesetval != NULL && + xpathObj->nodesetval->nodeNr != 0) + license->text = (gchar *)xmlNodeGetContent (xpathObj->nodesetval->nodeTab[0]); + xmlXPathFreeObject (xpathObj); + } + g_free (aux); + } + + /* Checking the license URI as defined by the Creative Commons + * Schema. This field is recomended to be checked by Creative + * Commons */ + xpathObj = xmlXPathEvalExpression (BAD_CAST LICENSE_URI, xpathCtx); + if (xpathObj != NULL) { + if (xpathObj->nodesetval != NULL && + xpathObj->nodesetval->nodeNr != 0) + license->uri = (gchar *)xmlNodeGetContent (xpathObj->nodesetval->nodeTab[0]); + xmlXPathFreeObject (xpathObj); + } + + /* Checking the web statement as defined by the XMP Rights + * Management Schema. Checking it out is a sort of above-and-beyond + * the basic recommendations by Creative Commons. It can be + * considered as a "reinforcement" approach to add certainty. */ + xpathObj = xmlXPathEvalExpression (BAD_CAST LICENSE_WEB_STATEMENT, xpathCtx); + if (xpathObj != NULL) { + if (xpathObj->nodesetval != NULL && + xpathObj->nodesetval->nodeNr != 0) + license->web_statement = (gchar *)xmlNodeGetContent (xpathObj->nodesetval->nodeTab[0]); + xmlXPathFreeObject (xpathObj); + } + } + xmlFree (marked); + + if (!license->text && !license->uri && !license->web_statement) { + ev_document_license_free (license); + return NULL; + } + + return license; +} + +static void +pdf_document_parse_metadata (const gchar *metadata, + EvDocumentInfo *info) +{ + xmlDocPtr doc; + xmlXPathContextPtr xpathCtx; + gchar *fmt; + + doc = xmlParseMemory (metadata, strlen (metadata)); + if (doc == NULL) + return; /* invalid xml metadata */ + + xpathCtx = xmlXPathNewContext (doc); + if (xpathCtx == NULL) { + xmlFreeDoc (doc); + return; /* invalid xpath context */ + } + + fmt = pdf_document_get_format_from_metadata (doc, xpathCtx); + if (fmt != NULL) { + g_free (info->format); + info->format = fmt; + } + + info->license = pdf_document_get_license_from_metadata (doc, xpathCtx); + + xmlXPathFreeContext (xpathCtx); + xmlFreeDoc (doc); +} + static EvDocumentInfo * pdf_document_get_info (EvDocument *document) @@ -567,7 +681,6 @@ pdf_document_get_info (EvDocument *document) PopplerPermissions permissions; EvPage *page; char *metadata; - char *fmt; info = g_new0 (EvDocumentInfo, 1); @@ -587,7 +700,8 @@ pdf_document_get_info (EvDocument *document) EV_DOCUMENT_INFO_LINEARIZED | EV_DOCUMENT_INFO_N_PAGES | EV_DOCUMENT_INFO_SECURITY | - EV_DOCUMENT_INFO_PAPER_SIZE; + EV_DOCUMENT_INFO_PAPER_SIZE | + EV_DOCUMENT_INFO_LICENSE; g_object_get (PDF_DOCUMENT (document)->document, "title", &(info->title), @@ -608,24 +722,16 @@ pdf_document_get_info (EvDocument *document) NULL); if (metadata != NULL) { - fmt = pdf_document_get_format_from_metadata(metadata); - if (fmt != NULL) { - g_free (info->format); - info->format = fmt; - } + pdf_document_parse_metadata (metadata, info); g_free (metadata); } info->n_pages = ev_document_get_n_pages (document); if (info->n_pages > 0) { - page = ev_document_get_page (document, 0); - ev_document_get_page_size (document, page, + ev_document_get_page_size (document, 0, &(info->paper_width), &(info->paper_height)); - g_object_unref (page); - - // Convert to mm. info->paper_width = info->paper_width / 72.0f * 25.4f; info->paper_height = info->paper_height / 72.0f * 25.4f; @@ -721,17 +827,42 @@ pdf_document_get_info (EvDocument *document) } static void -pdf_document_document_iface_init (EvDocumentIface *iface) -{ - iface->save = pdf_document_save; - iface->load = pdf_document_load; - iface->get_n_pages = pdf_document_get_n_pages; - iface->get_page = pdf_document_get_page; - iface->get_page_size = pdf_document_get_page_size; - iface->get_page_label = pdf_document_get_page_label; - iface->render = pdf_document_render; - iface->get_info = pdf_document_get_info; -}; +pdf_document_class_init (PdfDocumentClass *klass) +{ + GObjectClass *g_object_class = G_OBJECT_CLASS (klass); + EvDocumentClass *ev_document_class = EV_DOCUMENT_CLASS (klass); + + g_object_class->dispose = pdf_document_dispose; + + ev_document_class->save = pdf_document_save; + ev_document_class->load = pdf_document_load; + ev_document_class->get_n_pages = pdf_document_get_n_pages; + ev_document_class->get_page = pdf_document_get_page; + ev_document_class->get_page_size = pdf_document_get_page_size; + ev_document_class->get_page_label = pdf_document_get_page_label; + ev_document_class->render = pdf_document_render; + ev_document_class->get_info = pdf_document_get_info; +} + +/* EvDocumentSecurity */ +static gboolean +pdf_document_has_document_security (EvDocumentSecurity *document_security) +{ + /* FIXME: do we really need to have this? */ + return FALSE; +} + +static void +pdf_document_set_password (EvDocumentSecurity *document_security, + const char *password) +{ + PdfDocument *document = PDF_DOCUMENT (document_security); + + if (document->password) + g_free (document->password); + + document->password = g_strdup (password); +} static void pdf_document_security_iface_init (EvDocumentSecurityIface *iface)