path: root/mupdf/stream.c

#include <fitz.h>
#include <mupdf.h>

/*
 * Check if an object is a stream or not.
 */
int
pdf_isstream(pdf_xref *xref, int oid, int gen)
{
	fz_error *error;

	if (oid < 0 || oid >= xref->len)
		return 0;

	error = pdf_cacheobject(xref, oid, gen);
	if (error) {
		fz_warn("%s", error->msg);
		fz_droperror(error);
		return 0;
	}

	return xref->table[oid].stmbuf || xref->table[oid].stmofs;
}

/*
 * Create a filter given a name and param dictionary.
 */
static fz_error *
buildonefilter(fz_filter **fp, fz_obj *f, fz_obj *p)
{
	fz_filter *decompress;
	fz_filter *predict;
	fz_error *error;
	char *s;

	s = fz_toname(f);

	if (!strcmp(s, "ASCIIHexDecode") || !strcmp(s, "AHx"))
		return fz_newahxd(fp, p);

	if (!strcmp(s, "ASCII85Decode") || !strcmp(s, "A85"))
		return fz_newa85d(fp, p);

	if (!strcmp(s, "CCITTFaxDecode") || !strcmp(s, "CCF"))
		return fz_newfaxd(fp, p);

	if (!strcmp(s, "DCTDecode") || !strcmp(s, "DCT"))
		return fz_newdctd(fp, p);

	if (!strcmp(s, "RunLengthDecode") || !strcmp(s, "RL"))
		return fz_newrld(fp, p);

	if (!strcmp(s, "FlateDecode") || !strcmp(s, "Fl"))
	{
		if (fz_isdict(p))
		{
			fz_obj *obj = fz_dictgets(p, "Predictor");
			if (obj)
			{
				error = fz_newflated(&decompress, p);
				if (error)
					return error;

				error = fz_newpredictd(&predict, p);
				if (error)
				{
					fz_dropfilter(decompress);
					return error;
				}

				error = fz_newpipeline(fp, decompress, predict);
				fz_dropfilter(decompress);
				fz_dropfilter(predict);
				return error;
			}
		}
		return fz_newflated(fp, p);
	}

	if (!strcmp(s, "LZWDecode") || !strcmp(s, "LZW"))
	{
		if (fz_isdict(p))
		{
			fz_obj *obj = fz_dictgets(p, "Predictor");
			if (obj)
			{
				error = fz_newlzwd(&decompress, p);
				if (error)
					return error;

				error = fz_newpredictd(&predict, p);
				if (error)
				{
					fz_dropfilter(decompress);
					return error;
				}

				error = fz_newpipeline(fp, decompress, predict);
				fz_dropfilter(decompress);
				fz_dropfilter(predict);
				return error;
			}
		}
		return fz_newlzwd(fp, p);
	}

#ifdef HAVE_JBIG2DEC
	if (!strcmp(s, "JBIG2Decode"))
	{
		/* TODO: extract and feed JBIG2Global */
		return fz_newjbig2d(fp, p);
	}
#endif

#ifdef HAVE_JASPER
	if (!strcmp(s, "JPXDecode"))
		return fz_newjpxd(fp, p);
#endif

	return fz_throw("syntaxerror: unknown filter: %s", s);
}

/*
 * Build a chain of filters given filter names and param dicts.
 * If head is given, start filter chain with it.
 * Assume ownership of head.
 */
static fz_error *
buildfilterchain(fz_filter **filterp, fz_filter *head, fz_obj *fs, fz_obj *ps)
{
	fz_error *error;
	fz_filter *newhead;
	fz_filter *tail;
	fz_obj *f;
	fz_obj *p;
	int i;

	for (i = 0; i < fz_arraylen(fs); i++)
	{
		f = fz_arrayget(fs, i);
		if (fz_isarray(ps))
			p = fz_arrayget(ps, i);
		else
			p = nil;

		error = buildonefilter(&tail, f, p);
		if (error)
			return error;

		if (head)
		{
			error = fz_newpipeline(&newhead, head, tail);
			fz_dropfilter(head);
			fz_dropfilter(tail);
			if (error)
			{
				fz_dropfilter(newhead);
				return error;
			}
			head = newhead;
		}
		else
			head = tail;
	}

	*filterp = head;
	return nil;
}

/*
 * Build a filter for reading raw stream data.
 * This is a null filter to constrain reading to the
 * stream length, followed by a decryption filter.
 */
static fz_error *
makerawfilter(fz_filter **filterp, pdf_xref *xref, fz_obj *stmobj, int oid, int gen)
{
	fz_error *error;
	fz_filter *base;
	fz_obj *stmlen;
	int len;

	stmlen = fz_dictgets(stmobj, "Length");
	error = pdf_resolve(&stmlen, xref);
	if (error)
		return error;
	len = fz_toint(stmlen);
	fz_dropobj(stmlen);

	error = fz_newnullfilter(&base, len);
	if (error)
		return error;

	if (xref->crypt)
	{
		fz_filter *crypt;
		fz_filter *pipe;

		error = pdf_cryptstream(&crypt, xref->crypt, oid, gen);
		if (error)
		{
			fz_dropfilter(base);
			return error;
		}

		error = fz_newpipeline(&pipe, base, crypt);
		fz_dropfilter(base);
		fz_dropfilter(crypt);
		if (error)
			return error;

		*filterp = pipe;
	}
	else
	{
		*filterp = base;
	}

	return nil;
}

/*
 * Construct a filter to decode a stream, without
 * constraining to stream length, and without decryption.
 */
fz_error *
pdf_decodefilter(fz_filter **filterp, fz_obj *stmobj)
{
	fz_obj *filters;
	fz_obj *params;

	filters = fz_dictgetsa(stmobj, "Filter", "F");
	params = fz_dictgetsa(stmobj, "DecodeParms", "DP");

	if (filters)
	{
		if (fz_isname(filters))
			return buildonefilter(filterp, filters, params);
		else
			return buildfilterchain(filterp, nil, filters, params);
	}
	else
		return fz_newnullfilter(filterp, -1);

	return nil;
}

/*
 * Construct a filter to decode a stream, constraining
 * to stream length and decrypting.
 */
static fz_error *
makedecodefilter(fz_filter **filterp, pdf_xref *xref, fz_obj *stmobj, int oid, int gen)
{
	fz_error *error;
	fz_filter *base, *pipe, *tmp;
	fz_obj *filters;
	fz_obj *params;

	error = makerawfilter(&base, xref, stmobj, oid, gen);
	if (error)
		return error;

	filters = fz_dictgetsa(stmobj, "Filter", "F");
	params = fz_dictgetsa(stmobj, "DecodeParms", "DP");

	if (filters)
	{
		error = pdf_resolve(&filters, xref);
		if (error)
			goto cleanup0;

		if (params)
		{
			error = pdf_resolve(&params, xref);
			if (error)
				goto cleanup1;
		}

		if (fz_isname(filters))
		{
			error = buildonefilter(&tmp, filters, params);
			if (error)
				goto cleanup2;

			error = fz_newpipeline(&pipe, base, tmp);
			fz_dropfilter(tmp);
			if (error)
				goto cleanup2;
		}
		else
		{
			error = buildfilterchain(&pipe, base, filters, params);
			if (error)
				goto cleanup2;
		}

		if (params)
			fz_dropobj(params);

		fz_dropobj(filters);

		*filterp = pipe;
	}
	else
	{
		*filterp = base;
	}

	return nil;

cleanup2:
	if (params)
		fz_dropobj(params);
cleanup1:
	fz_dropobj(filters);
cleanup0:
	fz_dropfilter(base);
	return error;
}

/*
 * Open a stream for reading the raw (compressed but decrypted) data. 
 * Put the opened file in xref->stream. Using xref->file while this
 * is open is a bad idea.
 */
fz_error *
pdf_openrawstream(pdf_xref *xref, int oid, int gen)
{
	pdf_xrefentry *x;
	fz_error *error;
	fz_filter *filter;
	int n;

	if (oid < 0 || oid >= xref->len)
		return fz_throw("rangecheck: object id out of range");

	x = xref->table + oid;

	error = pdf_cacheobject(xref, oid, gen);
	if (error)
		return error;

	if (x->stmbuf)
	{
		return fz_openbuffer(&xref->stream, x->stmbuf, FZ_READ);
	}

	if (x->stmofs)
	{
		error = makerawfilter(&filter, xref, x->obj, oid, gen);
		if (error)
			return error;

		n = fz_seek(xref->file, x->stmofs, 0);
		if (n == -1)
		{
			fz_dropfilter(filter);
			return fz_ferror(xref->file);
		}

		error = fz_pushfilter(xref->file, filter);
		fz_dropfilter(filter);
		if (error)
			return error;

		xref->stream = xref->file;
		return nil;
	}

	return fz_throw("syntaxerror: object is not a stream");
}

/*
 * Open a stream for reading uncompressed data. 
 * Put the opened file in xref->stream.
 * Using xref->file while a stream is open is a Bad idea.
 */
fz_error *
pdf_openstream(pdf_xref *xref, int oid, int gen)
{
	pdf_xrefentry *x;
	fz_error *error;
	fz_filter *filter;
	int n;

	if (oid < 0 || oid >= xref->len)
		return fz_throw("rangecheck: object id out of range");

	x = xref->table + oid;

	error = pdf_cacheobject(xref, oid, gen);
	if (error)
		return error;

	if (x->stmbuf)
	{
		error = makedecodefilter(&filter, xref, x->obj, oid, gen);
		if (error)
			return error;

		error = fz_openbuffer(&xref->stream, x->stmbuf, FZ_READ);
		if (error)
		{
			fz_dropfilter(filter);
			return error;
		}

		error = fz_pushfilter(xref->stream, filter);
		fz_dropfilter(filter);
		return error;
	}

	if (x->stmofs)
	{
		error = makedecodefilter(&filter, xref, x->obj, oid, gen);
		if (error)
			return error;

		n = fz_seek(xref->file, x->stmofs, 0);
		if (n == -1)
		{
			fz_dropfilter(filter);
			return fz_ferror(xref->file);
		}

		error = fz_pushfilter(xref->file, filter);
		fz_dropfilter(filter);
		if (error)
			return error;

		xref->stream = xref->file;
		return nil;
	}

	return fz_throw("syntaxerror: object is not a stream");
}

/*
 * Close the xref->stream file opened by either
 * pdf_openrawstream or pdf_openstream.
 */
void
pdf_closestream(pdf_xref *xref)
{
	if (xref->stream == xref->file)
		fz_popfilter(xref->file);
	else
		fz_closefile(xref->stream);
	xref->stream = nil;
}

/*
 * Load raw (compressed but decrypted) contents of a stream into buf.
 */
fz_error *
pdf_loadrawstream(fz_buffer **bufp, pdf_xref *xref, int oid, int gen)
{
	fz_error *error;

	error = pdf_openrawstream(xref, oid, gen);
	if (error)
		return error;

	error = fz_readfile(bufp, xref->stream);

	pdf_closestream(xref);

	return error;
}

/*
 * Load uncompressed contents of a stream into buf.
 */
fz_error *
pdf_loadstream(fz_buffer **bufp, pdf_xref *xref, int oid, int gen)
{
	fz_error *error;

	error = pdf_openstream(xref, oid, gen);
	if (error)
		return error;

	error = fz_readfile(bufp, xref->stream);

	pdf_closestream(xref);

	return error;
}