From 7f18c9544d1bc707983005708cfa31ac169399ad Mon Sep 17 00:00:00 2001
From: Stepan Henek <stepan.henek@nic.cz>
Date: Thu, 27 Nov 2014 13:04:32 +0100
Subject: [PATCH] added option --raw-input which sets that the input file can
 be XML file (e.g. fodt)

---
 odt2txt.1 |  3 +++
 odt2txt.c | 39 +++++++++++++++++++++++++++++++++++----
 strbuf.c  | 40 ++++++++++++++++++++++++++++++++++++++++
 strbuf.h  |  6 ++++++
 4 files changed, 84 insertions(+), 4 deletions(-)

diff --git a/odt2txt.1 b/odt2txt.1
index 88caf12..9b1e5da 100644
--- a/odt2txt.1
+++ b/odt2txt.1
@@ -53,6 +53,9 @@ encoding will be used in automatic mode, use
 \fB\-\-raw\fR
 Print raw XML
 .TP
+\fB\-\-raw-input\fR
+Input file is a raw XML (fodt, fods, ...)
+.TP
 \fB\-\-version\fR
 Show version and copyright information
 .SH COPYRIGHT
diff --git a/odt2txt.c b/odt2txt.c
index 3480749..9b29fb7 100644
--- a/odt2txt.c
+++ b/odt2txt.c
@@ -45,6 +45,7 @@
 #define VERSION "0.5"
 
 static int opt_raw;
+static int opt_raw_input = 0;
 static char *opt_encoding;
 static int opt_width = 63;
 static const char *opt_filename;
@@ -138,6 +139,7 @@ static void usage(void)
 	       "Converts an OpenDocument or OpenOffice.org XML File to raw text.\n\n"
 	       "Syntax:   odt2txt [options] filename\n\n"
 	       "Options:  --raw         Print raw XML\n"
+	       "          --raw-input   Input file is a raw XML (fodt, fods, ...)\n"
 #ifdef NO_ICONV
 	       "          --encoding=X  Ignored. odt2txt has been built without iconv support.\n"
 	       "                        Output will always be encoded in UTF-8\n"
@@ -435,11 +437,34 @@ static STRBUF *read_from_zip(const char *zipfile, const char *filename)
 	return content;
 }
 
-static void format_doc(STRBUF *buf)
+static STRBUF *read_from_xml(const char *zipfile, const char *filename)
+{
+
+	FILE *in = fopen(zipfile, "rb");
+	if (in == 0) {
+		fprintf(stderr, "Can't open %s.\n", filename);
+		exit(EXIT_FAILURE);
+	}
+
+	STRBUF *content = strbuf_new();
+
+	strbuf_append_file(content, in);
+
+	fclose(in);
+
+	return content;
+}
+
+static void format_doc(STRBUF *buf, int raw_input)
 {
 	/* FIXME: Convert buffer to utf-8 first.  Are there
 	   OpenOffice texts which are not utf8-encoded? */
 
+	if (raw_input) {
+		RS_O(".*<office:body>", "<office:body>");                   /* only body */
+		RS_G("<office:binary-data>[^>]*</office:binary-data>", ""); /* remove binary */
+	}
+
 	/* headline, first level */
 	RS_E("<text:h[^>]*outline-level=\"1\"[^>]*>([^<]*)<[^>]*>", &h1);
 	RS_E("<text:h[^>]*>([^<]*)<[^>]*>", &h2);  /* other headlines */
@@ -451,7 +476,8 @@ static void format_doc(STRBUF *buf)
 	/* images */
 	RS_E("<draw:frame[^>]*draw:name=\"([^\"]*)\"[^>]*>", &image);
 
-	RS_G("<[^>]*>", ""); 	 /* replace all remaining tags */
+
+	RS_G("<[^>]*>", "");     /* replace all remaining tags */
 	RS_G("\n +", "\n");      /* remove indentations, e.g. kword */
 	RS_G("\n{3,}", "\n\n");  /* remove large vertical spaces */
 
@@ -480,6 +506,9 @@ int main(int argc, const char **argv)
 		if (!strcmp(argv[i], "--raw")) {
 			opt_raw = 1;
 			i++; continue;
+		} else if (!strcmp(argv[i], "--raw-input")) {
+			opt_raw_input = 1;
+			i++; continue;
 		} else if (!strncmp(argv[i], "--encoding=", 11)) {
 			size_t arglen = strlen(argv[i]) - 10;
 #ifdef iconvlist
@@ -563,11 +592,13 @@ int main(int argc, const char **argv)
 	}
 
 	/* read content.xml */
-	docbuf = read_from_zip(opt_filename, "content.xml");
+	docbuf = opt_raw_input ?
+		read_from_xml(opt_filename, "content.xml") :
+		read_from_zip(opt_filename, "content.xml");
 
 	if (!opt_raw) {
 		subst_doc(ic, docbuf);
-		format_doc(docbuf);
+		format_doc(docbuf, opt_raw_input);
 	}
 
 	wbuf = wrap(docbuf, opt_width);
diff --git a/strbuf.c b/strbuf.c
index 22b32c0..64985f9 100644
--- a/strbuf.c
+++ b/strbuf.c
@@ -156,6 +156,46 @@ int strbuf_subst(STRBUF *buf,
 	return diff;
 }
 
+size_t strbuf_append_file(STRBUF *buf, FILE *in)
+{
+	strbuf_check(buf);
+
+	/* save NULLOK flag */
+	int nullok = (buf->opt & STRBUF_NULLOK) ? 1 : 0;
+	strbuf_setopt(buf, STRBUF_NULLOK);
+
+
+	size_t len = 0;
+	size_t read_len = 0;
+	char readbuf[1024];
+	do {
+		read_len = fread(readbuf, 1, sizeof(readbuf), in);
+		len += read_len;
+
+		if (read_len > 0) {
+			while (buf->buf_sz < buf->len + sizeof(readbuf))
+				strbuf_grow(buf);
+
+			memcpy(buf->data + buf->len, readbuf, read_len);
+			buf->len += read_len;
+		}
+
+	} while (read_len == sizeof(readbuf));
+
+	/* terminate buffer */
+	if (buf->len + 1 > buf->buf_sz)
+		strbuf_grow(buf);
+	*(buf->data + buf->len) = '\0';
+
+	/* restore NULLOK option */
+	if (!nullok)
+		strbuf_unsetopt(buf, STRBUF_NULLOK);
+
+	strbuf_check(buf);
+
+	return len;
+}
+
 size_t strbuf_append_inflate(STRBUF *buf, FILE *in)
 {
 	size_t len;
diff --git a/strbuf.h b/strbuf.h
index 4e7500b..22f5c58 100644
--- a/strbuf.h
+++ b/strbuf.h
@@ -55,6 +55,12 @@ size_t strbuf_append(STRBUF *buf, const char *str);
  */
 size_t strbuf_append_inflate(STRBUF *buf, FILE *in);
 
+/*
+ * Reads a data stream from in and appends it to the buffer out.
+ * Returns the number of appended characters.
+ */
+size_t strbuf_append_file(STRBUF *buf, FILE *in);
+
 /*
  * Returns a pointer to the contained string.
  */