diff options
Diffstat (limited to 'pdf/pdf_page.c')
| -rw-r--r-- | pdf/pdf_page.c | 246 |
1 files changed, 246 insertions, 0 deletions
diff --git a/pdf/pdf_page.c b/pdf/pdf_page.c new file mode 100644 index 00000000..869f7a5c --- /dev/null +++ b/pdf/pdf_page.c @@ -0,0 +1,246 @@ +#include "fitz.h" +#include "mupdf.h" + +/* we need to combine all sub-streams into one for the content stream interpreter */ + +static fz_error +pdf_loadpagecontentsarray(fz_buffer **bigbufp, pdf_xref *xref, fz_obj *list) +{ + fz_error error; + fz_buffer *big; + fz_buffer *one; + int i; + + pdf_logpage("multiple content streams: %d\n", fz_arraylen(list)); + + /* TODO: openstream, read, close into big buffer at once */ + + big = fz_newbuffer(32 * 1024); + + for (i = 0; i < fz_arraylen(list); i++) + { + fz_obj *stm = fz_arrayget(list, i); + error = pdf_loadstream(&one, xref, fz_tonum(stm), fz_togen(stm)); + if (error) + { + fz_dropbuffer(big); + return fz_rethrow(error, "cannot load content stream part %d/%d (%d %d R)", i + 1, fz_arraylen(list), fz_tonum(stm), fz_togen(stm)); + } + + if (big->len + one->len + 1 > big->cap) + fz_resizebuffer(big, big->len + one->len + 1); + memcpy(big->data + big->len, one->data, one->len); + big->data[big->len + one->len] = ' '; + big->len += one->len + 1; + + fz_dropbuffer(one); + } + + *bigbufp = big; + return fz_okay; +} + +static fz_error +pdf_loadpagecontents(fz_buffer **bufp, pdf_xref *xref, fz_obj *obj) +{ + fz_error error; + + if (fz_isarray(obj)) + { + error = pdf_loadpagecontentsarray(bufp, xref, obj); + if (error) + return fz_rethrow(error, "cannot load content stream array (%d 0 R)", fz_tonum(obj)); + } + else if (pdf_isstream(xref, fz_tonum(obj), fz_togen(obj))) + { + error = pdf_loadstream(bufp, xref, fz_tonum(obj), fz_togen(obj)); + if (error) + return fz_rethrow(error, "cannot load content stream (%d 0 R)", fz_tonum(obj)); + } + else + { + fz_warn("page contents missing, leaving page blank"); + *bufp = fz_newbuffer(0); + } + + return fz_okay; +} + +/* We need to know whether to install a page-level transparency group */ + +static int pdf_resourcesuseblending(fz_obj *rdb); + +static int +pdf_extgstateusesblending(fz_obj *dict) +{ + fz_obj *obj; + + obj = fz_dictgets(dict, "BM"); + if (fz_isname(obj) && strcmp(fz_toname(obj), "Normal")) + return 1; + + return 0; +} + +static int +pdf_patternusesblending(fz_obj *dict) +{ + fz_obj *obj; + + obj = fz_dictgets(dict, "Resources"); + if (fz_isdict(obj) && pdf_resourcesuseblending(obj)) + return 1; + + obj = fz_dictgets(dict, "ExtGState"); + if (fz_isdict(obj) && pdf_extgstateusesblending(obj)) + return 1; + + return 0; +} + +static int +pdf_xobjectusesblending(fz_obj *dict) +{ + fz_obj *obj; + + obj = fz_dictgets(dict, "Resources"); + if (fz_isdict(obj) && pdf_resourcesuseblending(obj)) + return 1; + + return 0; +} + +static int +pdf_resourcesuseblending(fz_obj *rdb) +{ + fz_obj *dict; + fz_obj *tmp; + int i; + + /* stop on cyclic resource dependencies */ + if (fz_dictgets(rdb, ".useBM")) + return fz_tobool(fz_dictgets(rdb, ".useBM")); + + tmp = fz_newbool(0); + fz_dictputs(rdb, ".useBM", tmp); + fz_dropobj(tmp); + + dict = fz_dictgets(rdb, "ExtGState"); + for (i = 0; i < fz_dictlen(dict); i++) + if (pdf_extgstateusesblending(fz_dictgetval(dict, i))) + goto found; + + dict = fz_dictgets(rdb, "Pattern"); + for (i = 0; i < fz_dictlen(dict); i++) + if (pdf_patternusesblending(fz_dictgetval(dict, i))) + goto found; + + dict = fz_dictgets(rdb, "XObject"); + for (i = 0; i < fz_dictlen(dict); i++) + if (pdf_xobjectusesblending(fz_dictgetval(dict, i))) + goto found; + + return 0; + +found: + tmp = fz_newbool(1); + fz_dictputs(rdb, ".useBM", tmp); + fz_dropobj(tmp); + return 1; +} + +fz_error +pdf_loadpage(pdf_page **pagep, pdf_xref *xref, fz_obj *dict) +{ + fz_error error; + pdf_page *page; + fz_obj *obj; + fz_bbox bbox; + + pdf_logpage("load page {\n"); + + // TODO: move this to a more appropriate place + /* Ensure that we have a store for resource objects */ + if (!xref->store) + xref->store = pdf_newstore(); + + page = fz_malloc(sizeof(pdf_page)); + page->resources = nil; + page->contents = nil; + page->transparency = 0; + page->links = nil; + page->annots = nil; + + obj = fz_dictgets(dict, "MediaBox"); + bbox = fz_roundrect(pdf_torect(obj)); + if (fz_isemptyrect(pdf_torect(obj))) + { + fz_warn("cannot find page bounds, guessing page bounds."); + bbox.x0 = 0; + bbox.y0 = 0; + bbox.x1 = 612; + bbox.y1 = 792; + } + + obj = fz_dictgets(dict, "CropBox"); + if (fz_isarray(obj)) + { + fz_bbox cropbox = fz_roundrect(pdf_torect(obj)); + bbox = fz_intersectbbox(bbox, cropbox); + } + + page->mediabox.x0 = MIN(bbox.x0, bbox.x1); + page->mediabox.y0 = MIN(bbox.y0, bbox.y1); + page->mediabox.x1 = MAX(bbox.x0, bbox.x1); + page->mediabox.y1 = MAX(bbox.y0, bbox.y1); + + if (page->mediabox.x1 - page->mediabox.x0 < 1 || page->mediabox.y1 - page->mediabox.y0 < 1) + return fz_throw("invalid page size"); + + page->rotate = fz_toint(fz_dictgets(dict, "Rotate")); + + pdf_logpage("bbox [%d %d %d %d]\n", bbox.x0, bbox.y0, bbox.x1, bbox.y1); + pdf_logpage("rotate %d\n", page->rotate); + + obj = fz_dictgets(dict, "Annots"); + if (obj) + { + pdf_loadlinks(&page->links, xref, obj); + pdf_loadannots(&page->annots, xref, obj); + } + + page->resources = fz_dictgets(dict, "Resources"); + if (page->resources) + fz_keepobj(page->resources); + + obj = fz_dictgets(dict, "Contents"); + error = pdf_loadpagecontents(&page->contents, xref, obj); + if (error) + { + pdf_freepage(page); + return fz_rethrow(error, "cannot load page contents (%d %d R)", fz_tonum(obj), fz_togen(obj)); + } + + if (page->resources && pdf_resourcesuseblending(page->resources)) + page->transparency = 1; + + pdf_logpage("} %p\n", page); + + *pagep = page; + return fz_okay; +} + +void +pdf_freepage(pdf_page *page) +{ + pdf_logpage("drop page %p\n", page); + if (page->resources) + fz_dropobj(page->resources); + if (page->contents) + fz_dropbuffer(page->contents); + if (page->links) + pdf_freelink(page->links); + if (page->annots) + pdf_freeannot(page->annots); + fz_free(page); +} |