fz_error
pdf_cacheobject(pdf_xref *xref, int num, int gen)
{
fz_error error;
pdf_xrefentry *x;
int rnum, rgen;
if (num < 0 || num >= xref->len)
return fz_throw("object out of range (%d %d R); xref size %d", num, gen, xref->len);
x = &xref->table[num];
if (x->obj)
return fz_okay;
if (x->type == 'f')
{
x->obj = fz_newnull();
return fz_okay;
}
else if (x->type == 'n')
{
fz_seek(xref->file, x->ofs, 0);
error = pdf_parseindobj(&x->obj, xref, xref->file, xref->scratch, sizeof xref->scratch,
&rnum, &rgen, &x->stmofs);
if (error)
return fz_rethrow(error, "cannot parse object (%d %d R)", num, gen);
if (rnum != num)
return fz_throw("found object (%d %d R) instead of (%d %d R)", rnum, rgen, num, gen);
if (xref->crypt)
pdf_cryptobj(xref->crypt, x->obj, num, gen);
}
else if (x->type == 'o')
{
if (!x->obj)
{
error = pdf_loadobjstm(xref, x->ofs, 0, xref->scratch, sizeof xref->scratch);
if (error)
return fz_rethrow(error, "cannot load object stream containing object (%d %d R)", num, gen);
if (!x->obj)
return fz_throw("object (%d %d R) was not found in its object stream", num, gen);
}
}
else
{
return fz_throw("assert: corrupt xref struct");
}
return fz_okay;
}
static fz_error
pdf_readnewtrailer(pdf_xref *xref, char *buf, int cap)
{
fz_error error;
pdf_logxref("load new xref format trailer\n");
error = pdf_parseindobj(&xref->trailer, xref, xref->file, buf, cap, nil, nil, nil);
if (error)
return fz_rethrow(error, "cannot parse trailer (compressed)");
return fz_okay;
}
static fz_error
pdf_readnewxref(fz_obj **trailerp, pdf_xref *xref, char *buf, int cap)
{
fz_error error;
fz_stream *stm;
fz_obj *trailer;
fz_obj *index;
fz_obj *obj;
int num, gen, stmofs;
int size, w0, w1, w2;
int t;
int i;
pdf_logxref("load new xref format\n");
error = pdf_parseindobj(&trailer, xref, xref->file, buf, cap, &num, &gen, &stmofs);
if (error)
return fz_rethrow(error, "cannot parse compressed xref stream object");
obj = fz_dictgets(trailer, "Size");
if (!obj)
{
fz_dropobj(trailer);
return fz_throw("xref stream missing Size entry (%d %d R)", num, gen);
}
size = fz_toint(obj);
if (size >= xref->cap)
{
xref->cap = size + 1; /* for hack to allow broken pdf generators with off-by-one errors */
xref->table = fz_realloc(xref->table, xref->cap * sizeof(pdf_xrefentry));
}
if (size > xref->len)
{
for (i = xref->len; i < xref->cap; i++)
{
xref->table[i].ofs = 0;
xref->table[i].gen = 0;
xref->table[i].stmofs = 0;
xref->table[i].obj = nil;
xref->table[i].type = 0;
}
xref->len = size;
}
if (num < 0 || num >= xref->len)
{
if (num == xref->len && num < xref->cap)
{
/* allow broken pdf files that have off-by-one errors in the xref */
fz_warn("object id (%d %d R) out of range (0..%d)", num, gen, xref->len - 1);
xref->len ++;
}
else
{
fz_dropobj(trailer);
return fz_throw("object id (%d %d R) out of range (0..%d)", num, gen, xref->len - 1);
}
}
pdf_logxref("\tnum=%d gen=%d size=%d\n", num, gen, size);
obj = fz_dictgets(trailer, "W");
if (!obj) {
fz_dropobj(trailer);
return fz_throw("xref stream missing W entry (%d %d R)", num, gen);
}
w0 = fz_toint(fz_arrayget(obj, 0));
w1 = fz_toint(fz_arrayget(obj, 1));
w2 = fz_toint(fz_arrayget(obj, 2));
index = fz_dictgets(trailer, "Index");
error = pdf_openstreamat(&stm, xref, num, gen, trailer, stmofs);
if (error)
{
fz_dropobj(trailer);
return fz_rethrow(error, "cannot open compressed xref stream (%d %d R)", num, gen);
}
if (!index)
{
error = pdf_readnewxrefsection(xref, stm, 0, size, w0, w1, w2);
if (error)
{
fz_close(stm);
fz_dropobj(trailer);
return fz_rethrow(error, "cannot read xref stream (%d %d R)", num, gen);
}
}
else
{
for (t = 0; t < fz_arraylen(index); t += 2)
{
int i0 = fz_toint(fz_arrayget(index, t + 0));
int i1 = fz_toint(fz_arrayget(index, t + 1));
error = pdf_readnewxrefsection(xref, stm, i0, i1, w0, w1, w2);
if (error)
{
fz_close(stm);
fz_dropobj(trailer);
return fz_rethrow(error, "cannot read xref stream section (%d %d R)", num, gen);
}
}
}
fz_close(stm);
*trailerp = trailer;
return fz_okay;
}
static fz_error
readnewxref(fz_obj **trailerp, pdf_xref *xref, char *buf, int cap)
{
fz_error error;
fz_stream *stm;
fz_obj *trailer;
fz_obj *index;
fz_obj *obj;
int oid, gen, stmofs;
int size, w0, w1, w2;
int t;
pdf_logxref("load new xref format\n");
error = pdf_parseindobj(&trailer, xref, xref->file, buf, cap, &oid, &gen, &stmofs);
if (error)
return fz_rethrow(error, "cannot parse compressed xref stream object");
if (oid < 0 || oid >= xref->len)
{
if (oid == xref->len && oid < xref->cap)
{
/* allow broken pdf files that have off-by-one errors in the xref */
fz_warn("object id (%d %d R) out of range (0..%d)", oid, gen, xref->len - 1);
xref->len ++;
}
else
{
fz_dropobj(trailer);
return fz_throw("object id (%d %d R) out of range (0..%d)", oid, gen, xref->len - 1);
}
}
xref->table[oid].type = 'n';
xref->table[oid].gen = gen;
xref->table[oid].obj = fz_keepobj(trailer);
xref->table[oid].stmofs = stmofs;
xref->table[oid].ofs = 0;
obj = fz_dictgets(trailer, "Size");
if (!obj)
{
fz_dropobj(trailer);
return fz_throw("xref stream missing Size entry");
}
size = fz_toint(obj);
obj = fz_dictgets(trailer, "W");
if (!obj) {
fz_dropobj(trailer);
return fz_throw("xref stream missing W entry");
}
w0 = fz_toint(fz_arrayget(obj, 0));
w1 = fz_toint(fz_arrayget(obj, 1));
w2 = fz_toint(fz_arrayget(obj, 2));
index = fz_dictgets(trailer, "Index");
error = pdf_openstream(&stm, xref, oid, gen);
if (error)
{
fz_dropobj(trailer);
return fz_rethrow(error, "cannot open compressed xref stream");
}
if (!index)
{
error = readnewxrefsection(xref, stm, 0, size, w0, w1, w2);
if (error)
{
fz_dropstream(stm);
fz_dropobj(trailer);
return fz_rethrow(error, "cannot read xref stream");
}
}
else
{
for (t = 0; t < fz_arraylen(index); t += 2)
{
int i0 = fz_toint(fz_arrayget(index, t + 0));
int i1 = fz_toint(fz_arrayget(index, t + 1));
error = readnewxrefsection(xref, stm, i0, i1, w0, w1, w2);
if (error)
{
fz_dropstream(stm);
fz_dropobj(trailer);
return fz_rethrow(error, "cannot read xref stream section");
}
}
}
fz_dropstream(stm);
*trailerp = trailer;
return fz_okay;
}
