int
newPdfFile(
soPdfFile* pdfFile
)
{
fz_error *error;
assert(pdfFile != NULL);
error = pdf_newxref(&pdfFile->xref);
if (error)
return soPdfError(error);
error = pdf_initxref(pdfFile->xref);
if (error)
return soPdfError(error);
error = fz_newarray(&pdfFile->pagelist, 100);
if (error)
return soPdfError(error);
error = fz_newarray(&pdfFile->editobjs, 100);
if (error)
return soPdfError(error);
return 0;
}
static void addhexfilter(fz_obj *dict)
{
fz_obj *f, *dp, *newf, *newdp;
fz_obj *ahx, *nullobj;
ahx = fz_newname("ASCIIHexDecode");
nullobj = fz_newnull();
newf = newdp = nil;
f = fz_dictgets(dict, "Filter");
dp = fz_dictgets(dict, "DecodeParms");
if (fz_isname(f))
{
newf = fz_newarray(2);
fz_arraypush(newf, ahx);
fz_arraypush(newf, f);
f = newf;
if (fz_isdict(dp))
{
newdp = fz_newarray(2);
fz_arraypush(newdp, nullobj);
fz_arraypush(newdp, dp);
dp = newdp;
}
}
else if (fz_isarray(f))
{
fz_arrayinsert(f, ahx);
if (fz_isarray(dp))
fz_arrayinsert(dp, nullobj);
}
else
f = ahx;
fz_dictputs(dict, "Filter", f);
if (dp)
fz_dictputs(dict, "DecodeParms", dp);
fz_dropobj(ahx);
fz_dropobj(nullobj);
if (newf)
fz_dropobj(newf);
if (newdp)
fz_dropobj(newdp);
}
static fz_error parsearray(fz_obj **obj, pdf_xref *xref, char **sp, struct vap *v)
{
fz_error error;
fz_obj *a;
fz_obj *o;
char *s = *sp;
error = fz_newarray(&a, 8);
if (error)
return fz_rethrow(error, "cannot create array");
s ++; /* skip '[' */
while (*s)
{
skipwhite(&s);
if (*s == ']')
{
s ++;
break;
}
error = parseobj(&o, xref, &s, v);
if (error)
{
fz_dropobj(a);
return fz_rethrow(error, "cannot parse item");
}
error = fz_arraypush(a, o);
if (error)
{
fz_dropobj(o);
fz_dropobj(a);
return fz_rethrow(error, "cannot add item to array");
}
fz_dropobj(o);
}
*obj = a;
*sp = s;
return fz_okay;
}
static void retainpages(int argc, char **argv)
{
fz_error error;
fz_obj *oldroot, *root, *pages, *kids, *countobj, *parent;
/* Load the old page tree */
error = pdf_loadpagetree(xref);
if (error)
die(fz_rethrow(error, "cannot load page tree"));
/* Keep only pages/type entry to avoid references to unretained pages */
oldroot = fz_dictgets(xref->trailer, "Root");
pages = fz_dictgets(oldroot, "Pages");
root = fz_newdict(2);
fz_dictputs(root, "Type", fz_dictgets(oldroot, "Type"));
fz_dictputs(root, "Pages", fz_dictgets(oldroot, "Pages"));
pdf_updateobject(xref, fz_tonum(oldroot), fz_togen(oldroot), root);
fz_dropobj(root);
/* Create a new kids array with only the pages we want to keep */
parent = fz_newindirect(fz_tonum(pages), fz_togen(pages), xref);
kids = fz_newarray(1);
/* Retain pages specified */
while (argc - fz_optind)
{
int page, spage, epage;
char *spec, *dash;
char *pagelist = argv[fz_optind];
spec = fz_strsep(&pagelist, ",");
while (spec)
{
dash = strchr(spec, '-');
if (dash == spec)
spage = epage = pdf_getpagecount(xref);
else
spage = epage = atoi(spec);
if (dash)
{
if (strlen(dash) > 1)
epage = atoi(dash + 1);
else
epage = pdf_getpagecount(xref);
}
if (spage > epage)
page = spage, spage = epage, epage = page;
if (spage < 1)
spage = 1;
if (epage > pdf_getpagecount(xref))
epage = pdf_getpagecount(xref);
for (page = spage; page <= epage; page++)
{
fz_obj *pageobj = pdf_getpageobject(xref, page);
fz_obj *pageref = pdf_getpageref(xref, page);
fz_dictputs(pageobj, "Parent", parent);
/* Store page object in new kids array */
fz_arraypush(kids, pageref);
}
spec = fz_strsep(&pagelist, ",");
}
fz_optind++;
}
fz_dropobj(parent);
/* Update page count and kids array */
countobj = fz_newint(fz_arraylen(kids));
fz_dictputs(pages, "Count", countobj);
fz_dropobj(countobj);
fz_dictputs(pages, "Kids", kids);
fz_dropobj(kids);
}
fz_error
pdf_parsearray(fz_obj **op, pdf_xref *xref, fz_stream *file, char *buf, int cap)
{
fz_error error = fz_okay;
fz_obj *ary = nil;
fz_obj *obj = nil;
int a = 0, b = 0, n = 0;
pdf_token_e tok;
int len;
ary = fz_newarray(4);
while (1)
{
error = pdf_lex(&tok, file, buf, cap, &len);
if (error)
{
fz_dropobj(ary);
return fz_rethrow(error, "cannot parse array");
}
if (tok != PDF_TINT && tok != PDF_TR)
{
if (n > 0)
{
obj = fz_newint(a);
fz_arraypush(ary, obj);
fz_dropobj(obj);
}
if (n > 1)
{
obj = fz_newint(b);
fz_arraypush(ary, obj);
fz_dropobj(obj);
}
n = 0;
}
if (tok == PDF_TINT && n == 2)
{
obj = fz_newint(a);
fz_arraypush(ary, obj);
fz_dropobj(obj);
a = b;
n --;
}
switch (tok)
{
case PDF_TCARRAY:
*op = ary;
return fz_okay;
case PDF_TINT:
if (n == 0)
a = atoi(buf);
if (n == 1)
b = atoi(buf);
n ++;
break;
case PDF_TR:
if (n != 2)
{
fz_dropobj(ary);
return fz_throw("cannot parse indirect reference in array");
}
obj = fz_newindirect(a, b, xref);
fz_arraypush(ary, obj);
fz_dropobj(obj);
n = 0;
break;
case PDF_TOARRAY:
error = pdf_parsearray(&obj, xref, file, buf, cap);
if (error)
{
fz_dropobj(ary);
return fz_rethrow(error, "cannot parse array");
}
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TODICT:
error = pdf_parsedict(&obj, xref, file, buf, cap);
if (error)
{
fz_dropobj(ary);
return fz_rethrow(error, "cannot parse array");
}
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TNAME:
obj = fz_newname(buf);
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TREAL:
obj = fz_newreal(atof(buf));
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TSTRING:
obj = fz_newstring(buf, len);
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TTRUE:
obj = fz_newbool(1);
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TFALSE:
obj = fz_newbool(0);
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
case PDF_TNULL:
obj = fz_newnull();
fz_arraypush(ary, obj);
fz_dropobj(obj);
break;
default:
fz_dropobj(ary);
return fz_throw("cannot parse token in array");
}
}
}
void
editmain(int argc, char **argv)
{
char *outfile = "out.pdf";
fz_error *error;
int c;
while ((c = getopt(argc, argv, "o:")) != -1)
{
switch (c)
{
case 'o':
outfile = optarg;
break;
default:
editusage();
break;
}
}
if (optind == argc)
editusage();
error = pdf_newxref(&editxref);
if (error)
die(error);
error = pdf_initxref(editxref);
if (error)
die(error);
error = fz_newarray(&editpagelist, 100);
if (error)
die(error);
error = fz_newarray(&editmodelist, 100);
if (error)
die(error);
while (optind < argc)
{
if (strstr(argv[optind], ".pdf"))
{
if (editobjects)
editflushobjects();
opensrc(argv[optind], "", 1);
error = fz_newarray(&editobjects, 100);
if (error)
die(error);
}
else if (!strcmp(argv[optind], "copy"))
editmode = COPY;
else if (!strcmp(argv[optind], "over"))
editmode = OVER;
else if (!strcmp(argv[optind], "2up"))
editmode = NUP2;
else if (!strcmp(argv[optind], "4up"))
editmode = NUP4;
else if (!strcmp(argv[optind], "8up"))
editmode = NUP8;
else
editpages(argv[optind]);
optind++;
}
if (editobjects)
editflushobjects();
closesrc();
editflushpagetree();
editflushcatalog();
error = pdf_savexref(editxref, outfile, nil);
if (error)
die(error);
pdf_closexref(editxref);
}
fz_error *
pdf_parsearray(fz_obj **op, fz_stream *file, char *buf, int cap)
{
fz_error *error = nil;
fz_obj *ary = nil;
fz_obj *obj = nil;
int a = 0, b = 0, n = 0;
int tok, len;
error = fz_newarray(op, 4);
if (error) return error;
ary = *op;
while (1)
{
tok = pdf_lex(file, buf, cap, &len);
if (tok != PDF_TINT && tok != PDF_TR)
{
if (n > 0)
{
error = fz_newint(&obj, a);
if (error) goto cleanup;
error = fz_arraypush(ary, obj);
if (error) goto cleanup;
fz_dropobj(obj);
obj = nil;
}
if (n > 1)
{
error = fz_newint(&obj, b);
if (error) goto cleanup;
error = fz_arraypush(ary, obj);
if (error) goto cleanup;
fz_dropobj(obj);
obj = nil;
}
n = 0;
}
if (tok == PDF_TINT && n == 2)
{
error = fz_newint(&obj, a);
if (error) goto cleanup;
error = fz_arraypush(ary, obj);
if (error) goto cleanup;
fz_dropobj(obj);
obj = nil;
a = b;
n --;
}
switch (tok)
{
case PDF_TCARRAY:
return nil;
case PDF_TINT:
if (n == 0)
a = atoi(buf);
if (n == 1)
b = atoi(buf);
n ++;
break;
case PDF_TR:
if (n != 2)
goto cleanup;
error = fz_newindirect(&obj, a, b);
if (error) goto cleanup;
n = 0;
break;
case PDF_TOARRAY: error = pdf_parsearray(&obj, file, buf, cap); break;
case PDF_TODICT: error = pdf_parsedict(&obj, file, buf, cap); break;
case PDF_TNAME: error = fz_newname(&obj, buf); break;
case PDF_TREAL: error = fz_newreal(&obj, atof(buf)); break;
case PDF_TSTRING: error = fz_newstring(&obj, buf, len); break;
case PDF_TTRUE: error = fz_newbool(&obj, 1); break;
case PDF_TFALSE: error = fz_newbool(&obj, 0); break;
case PDF_TNULL: error = fz_newnull(&obj); break;
default: goto cleanup;
}
if (error) goto cleanup;
if (obj)
{
error = fz_arraypush(ary, obj);
if (error) goto cleanup;
fz_dropobj(obj);
}
obj = nil;
}
cleanup:
if (obj) fz_dropobj(obj);
if (ary) fz_dropobj(ary);
if (error) return error;
return fz_throw("syntaxerror: corrupt array");
}
