/* -*- Mode: C; tab-width: 8; indent-tabs-mode: t; c-basic-offset: 8 -*- */
/*
 * soup-content-sniffer.c
 *
 * Copyright (C) 2009, 2013 Gustavo Noronha Silva.
 *
 * This code implements the following specification:
 *
 *  http://mimesniff.spec.whatwg.org/ as of 11 June 2013
 */

#ifdef HAVE_CONFIG_H
#include <config.h>
#endif

#include <string.h>

#include "soup-content-sniffer.h"
#include "soup.h"
#include "soup-content-processor.h"
#include "soup-content-sniffer-stream.h"
#include "soup-message-private.h"

/**
 * SECTION:soup-content-sniffer
 * @short_description: Content sniffing for SoupSession
 *
 * A #SoupContentSniffer tries to detect the actual content type of
 * the files that are being downloaded by looking at some of the data
 * before the #SoupMessage emits its #SoupMessage::got-headers signal.
 * #SoupContentSniffer implements #SoupSessionFeature, so you can add
 * content sniffing to a session with soup_session_add_feature() or
 * soup_session_add_feature_by_type().
 *
 * Since: 2.28
 **/

static void soup_content_sniffer_session_feature_init (SoupSessionFeatureInterface *feature_interface, gpointer interface_data);

static SoupContentProcessorInterface *soup_content_sniffer_default_content_processor_interface;
static void soup_content_sniffer_content_processor_init (SoupContentProcessorInterface *interface, gpointer interface_data);


G_DEFINE_TYPE_WITH_CODE (SoupContentSniffer, soup_content_sniffer, G_TYPE_OBJECT,
			 G_IMPLEMENT_INTERFACE (SOUP_TYPE_SESSION_FEATURE,
						soup_content_sniffer_session_feature_init)
			 G_IMPLEMENT_INTERFACE (SOUP_TYPE_CONTENT_PROCESSOR,
						soup_content_sniffer_content_processor_init))


static GInputStream *
soup_content_sniffer_content_processor_wrap_input (SoupContentProcessor *processor,
						   GInputStream *base_stream,
						   SoupMessage *msg,
						   GError **error)
{
	return g_object_new (SOUP_TYPE_CONTENT_SNIFFER_STREAM,
			     "base-stream", base_stream,
			     "message", msg,
			     "sniffer", SOUP_CONTENT_SNIFFER (processor),
			     NULL);
}

static void
soup_content_sniffer_content_processor_init (SoupContentProcessorInterface *processor_interface,
                                            gpointer interface_data)
{
	soup_content_sniffer_default_content_processor_interface =
		g_type_default_interface_peek (SOUP_TYPE_CONTENT_PROCESSOR);

	processor_interface->processing_stage = SOUP_STAGE_BODY_DATA;
	processor_interface->wrap_input = soup_content_sniffer_content_processor_wrap_input;
}

static void
soup_content_sniffer_init (SoupContentSniffer *content_sniffer)
{
}

/* This table is based on the HTML5 spec;
 * See 2.7.4 Content-Type sniffing: unknown type
 */
typedef struct {
	/* @has_ws is TRUE if @pattern contains "generic" whitespace */
	gboolean      has_ws;
	const guchar *mask;
	const guchar *pattern;
	guint         pattern_length;
	const char   *sniffed_type;
	gboolean      scriptable;
} SoupContentSnifferPattern;

static SoupContentSnifferPattern types_table[] = {
	{ FALSE,
	  (const guchar *)"\xFF\xFF\xDF\xDF\xDF\xDF\xDF\xDF\xDF\xFF\xDF\xDF\xDF\xDF",
	  (const guchar *)"\x3C\x21\x44\x4F\x43\x54\x59\x50\x45\x20\x48\x54\x4D\x4C",
	  14,
	  "text/html",
	  TRUE },

	{ TRUE,
	  (const guchar *)"\xFF\xFF\xDF\xDF\xDF\xDF",
	  (const guchar *)" \x3C\x48\x54\x4D\x4C",
	  5,
	  "text/html",
	  TRUE },

	{ TRUE,
	  (const guchar *)"\xFF\xFF\xDF\xDF\xDF\xDF",
	  (const guchar *)" \x3C\x48\x45\x41\x44",
	  5,
	  "text/html",
	  TRUE },

	{ TRUE,
	  (const guchar *)"\xFF\xFF\xDF\xDF\xDF\xDF\xDF\xDF",
	  (const guchar *)" \x3C\x53\x43\x52\x49\x50\x54",
	  7,
	  "text/html",
	  TRUE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\xFF\xFF",
	  (const guchar *)"\x25\x50\x44\x46\x2D",
	  5,
	  "application/pdf",
	  TRUE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF",
	  (const guchar *)"\x25\x21\x50\x53\x2D\x41\x64\x6F\x62\x65\x2D",
	  11,
	  "application/postscript",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\x00\x00",
	  (const guchar *)"\xFE\xFF\x00\x00",
	  4,
	  "text/plain",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\x00\x00",
	  (const guchar *)"\xFF\xFF\x00\x00",
	  4,
	  "text/plain",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\x00",
	  (const guchar *)"\xEF\xBB\xBF\x00",
	  4,
	  "text/plain",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\xFF\xFF\xFF",
	  (const guchar *)"\x47\x49\x46\x38\x37\x61",
	  6,
	  "image/gif",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\xFF\xFF\xFF",
	  (const guchar *)"\x47\x49\x46\x38\x39\x61",
	  6,
	  "image/gif",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF",
	  (const guchar *)"\x89\x50\x4E\x47\x0D\x0A\x1A\x0A",
	  8,
	  "image/png",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF",
	  (const guchar *)"\xFF\xD8\xFF",
	  3,
	  "image/jpeg",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF",
	  (const guchar *)"\x42\x4D",
	  2,
	  "image/bmp",
	  FALSE },

	{ FALSE,
	  (const guchar *)"\xFF\xFF\xFF\xFF",
	  (const guchar *)"\x00\x00\x01\x00",
	  4,
	  "image/vnd.microsoft.icon",
	  FALSE }
};

/* Whether a given byte looks like it might be part of binary content.
 * Source: HTML5 spec; borrowed from the Chromium mime sniffer code,
 * which is BSD-licensed
 */
static char byte_looks_binary[] = {
	1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1, 0, 0, 1, 1,  /* 0x00 - 0x0F */
	1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1,  /* 0x10 - 0x1F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x20 - 0x2F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x30 - 0x3F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x40 - 0x4F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x50 - 0x5F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x60 - 0x6F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x70 - 0x7F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x80 - 0x8F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x90 - 0x9F */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0xA0 - 0xAF */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0xB0 - 0xBF */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0xC0 - 0xCF */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0xD0 - 0xDF */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0xE0 - 0xEF */
	0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0xF0 - 0xFF */
};

/* HTML5: 2.7.4 Content-Type sniffing: unknown type */
static char*
sniff_unknown (SoupContentSniffer *sniffer, SoupBuffer *buffer,
	       gboolean for_text_or_binary)
{
	const guchar *resource = (const guchar *)buffer->data;
	int resource_length = MIN (512, buffer->length);
	int i;

	for (i = 0; i < G_N_ELEMENTS (types_table); i++) {
		SoupContentSnifferPattern *type_row = &(types_table[i]);

		/* The scriptable types should be skiped for the text
		 * or binary path, but considered for other paths */
		if (for_text_or_binary && type_row->scriptable)
			continue;

		if (type_row->has_ws) {
			int index_stream = 0;
			int index_pattern = 0;
			gboolean skip_row = FALSE;

			while ((index_stream < resource_length) &&
			       (index_pattern <= type_row->pattern_length)) {
				/* Skip insignificant white space ("WS" in the spec) */
				if (type_row->pattern[index_pattern] == ' ') {
					if (resource[index_stream] == '\x09' ||
					    resource[index_stream] == '\x0a' ||
					    resource[index_stream] == '\x0c' ||
					    resource[index_stream] == '\x0d' ||
					    resource[index_stream] == '\x20')
						index_stream++;
					else
						index_pattern++;
				} else {
					if ((type_row->mask[index_pattern] & resource[index_stream]) != type_row->pattern[index_pattern]) {
						skip_row = TRUE;
						break;
					}
					index_pattern++;
					index_stream++;
				}
			}

			if (skip_row)
				continue;

			if (index_pattern > type_row->pattern_length)
				return g_strdup (type_row->sniffed_type);
		} else {
			int j;

			if (resource_length < type_row->pattern_length)
				continue;

			for (j = 0; j < type_row->pattern_length; j++) {
				if ((type_row->mask[j] & resource[j]) != type_row->pattern[j])
					break;
			}

			/* This means our comparison above matched completely */
			if (j == type_row->pattern_length)
				return g_strdup (type_row->sniffed_type);
		}
	}

	if (for_text_or_binary)
		return g_strdup ("application/octet-stream");

	for (i = 0; i < resource_length; i++) {
		if (byte_looks_binary[resource[i]])
			return g_strdup ("application/octet-stream");
	}

	return g_strdup ("text/plain");
}

/* MIMESNIFF: 7.2 Sniffing a mislabeled binary resource */
static char*
sniff_text_or_binary (SoupContentSniffer *sniffer, SoupBuffer *buffer)
{
	const guchar *resource = (const guchar *)buffer->data;
	int resource_length = MIN (512, buffer->length);
	gboolean looks_binary = FALSE;
	int i;

	/* 2. Detecting UTF-16BE, UTF-16LE BOMs means it's text/plain */
	if (resource_length >= 2) {
		if ((resource[0] == 0xFE && resource[1] == 0xFF) ||
		    (resource[0] == 0xFF && resource[1] == 0xFE))
			return g_strdup ("text/plain");
	}

	/* 3. UTF-8 BOM. */
	if (resource_length >= 3) {
		if (resource[0] == 0xEF && resource[1] == 0xBB && resource[2] == 0xBF)
			return g_strdup ("text/plain");
	}

	/* 4. Look to see if any of the first n bytes looks binary */
	for (i = 0; i < resource_length; i++) {
		if (byte_looks_binary[resource[i]]) {
			looks_binary = TRUE;
			break;
		}
	}

	if (!looks_binary)
		return g_strdup ("text/plain");

	/* 5. Execute 7.1 Identifying a resource with an unknown MIME type.
	 * TODO: sniff-scriptable needs to be unset.
	 */
	return sniff_unknown (sniffer, buffer, TRUE);
}

static char*
sniff_images (SoupContentSniffer *sniffer, SoupBuffer *buffer,
	      const char *content_type)
{
	const guchar *resource = (const guchar *)buffer->data;
	int resource_length = MIN (512, buffer->length);
	int i;

	for (i = 0; i < G_N_ELEMENTS (types_table); i++) {
		SoupContentSnifferPattern *type_row = &(types_table[i]);

		if (resource_length < type_row->pattern_length)
			continue;

		if (!g_str_has_prefix (type_row->sniffed_type, "image/"))
			continue;

		/* All of the image types use all-\xFF for the mask,
		 * so we can just memcmp.
		 */
		if (memcmp (type_row->pattern, resource, type_row->pattern_length) == 0)
			return g_strdup (type_row->sniffed_type);
	}

	return g_strdup (content_type);
}

static char*
sniff_feed_or_html (SoupContentSniffer *sniffer, SoupBuffer *buffer)
{
	const guchar *resource = (const guchar *)buffer->data;
	int resource_length = MIN (512, buffer->length);
	int pos = 0;

	if (resource_length < 3)
		goto text_html;

	/* Skip a leading UTF-8 BOM */
	if (resource[0] == 0xEF && resource[1] == 0xBB && resource[2] == 0xBF)
		pos = 3;

 look_for_tag:
	if (pos > resource_length)
		goto text_html;

	/* Skip insignificant white space */
	while ((resource[pos] == '\x09') ||
	       (resource[pos] == '\x20') ||
	       (resource[pos] == '\x0A') ||
	       (resource[pos] == '\x0D')) {
		pos++;

		if (pos > resource_length)
			goto text_html;
	}

	/* != < */
	if (resource[pos] != '\x3C')
		return g_strdup ("text/html");

	pos++;

	if ((pos + 2) > resource_length)
		goto text_html;

	/* Skipping comments */
	if ((resource[pos] == '\x2D') ||
	    (resource[pos+1] == '\x2D') ||
	    (resource[pos+2] == '\x3E')) {
		pos = pos + 3;

		if ((pos + 2) > resource_length)
			goto text_html;

		while ((resource[pos] != '\x2D') &&
		       (resource[pos+1] != '\x2D') &&
		       (resource[pos+2] != '\x3E')) {
			pos++;

			if ((pos + 2) > resource_length)
				goto text_html;
		}

		goto look_for_tag;
	}

	if (pos > resource_length)
		goto text_html;

	/* == ! */
	if (resource[pos] == '\x21') {
		do {
			pos++;

			if (pos > resource_length)
				goto text_html;
		} while (resource[pos] != '\x3E');

		pos++;

		goto look_for_tag;
	} else if (resource[pos] == '\x3F') { /* ? */
		do {
			pos++;

			if ((pos + 1) > resource_length)
				goto text_html;
		} while ((resource[pos] != '\x3F') &&
			 (resource[pos+1] != '\x3E'));

		pos = pos + 2;

		goto look_for_tag;
	}

	if ((pos + 2) > resource_length)
		goto text_html;

	if ((resource[pos] == '\x72') &&
	    (resource[pos+1] == '\x73') &&
	    (resource[pos+2] == '\x73'))
		return g_strdup ("application/rss+xml");

	if ((pos + 3) > resource_length)
		goto text_html;

	if ((resource[pos] == '\x66') &&
	    (resource[pos+1] == '\x65') &&
	    (resource[pos+2] == '\x65') &&
	    (resource[pos+3] == '\x64'))
		return g_strdup ("application/atom+xml");

 text_html:
	return g_strdup ("text/html");
}

static char *
soup_content_sniffer_real_sniff (SoupContentSniffer *sniffer, SoupMessage *msg,
				 SoupBuffer *buffer, GHashTable **params)
{
	const char *content_type;
	const char *x_content_type_options;

	content_type = soup_message_headers_get_content_type (msg->response_headers, params);

	/* MIMESNIFF: 7 Determining the sniffed MIME type of a resource. */

	/* 1. Unknown/undefined supplied type respecting sniff-scritable. */
	if ((content_type == NULL) ||
	    !g_ascii_strcasecmp (content_type, "unknown/unknown") ||
	    !g_ascii_strcasecmp (content_type, "application/unknown") ||
	    !g_ascii_strcasecmp (content_type, "*/*"))
		return sniff_unknown (sniffer, buffer, FALSE);

	/* 2. If nosniff is specified in X-Content-Type-Options use the supplied MIME type. */
	x_content_type_options = soup_message_headers_get_one (msg->response_headers, "X-Content-Type-Options");
	if (!g_strcmp0 (x_content_type_options, "nosniff"))
		return g_strdup (content_type);

	/* 3. check-for-apache-bug */
	if ((content_type != NULL) &&
	    (g_str_equal (content_type, "text/plain") ||
	     g_str_equal (content_type, "text/plain; charset=ISO-8859-1") ||
	     g_str_equal (content_type, "text/plain; charset=iso-8859-1") ||
	     g_str_equal (content_type, "text/plain; charset=UTF-8")))
		return sniff_text_or_binary (sniffer, buffer);

	if (g_str_has_suffix (content_type, "+xml") ||
	    !g_ascii_strcasecmp (content_type, "text/xml") ||
	    !g_ascii_strcasecmp (content_type, "application/xml"))
		return g_strdup (content_type);

	/* 2.7.5 Content-Type sniffing: image
	 * The spec says:
	 *
	 *   If the resource's official type is "image/svg+xml", then
	 *   the sniffed type of the resource is its official type (an
	 *   XML type)
	 *
	 * The XML case is handled by the if above; if you refactor
	 * this code, keep this in mind.
	 */
	if (!g_ascii_strncasecmp (content_type, "image/", 6))
		return sniff_images (sniffer, buffer, content_type);

	/* If we got text/plain, use text_or_binary */
	if (g_str_equal (content_type, "text/plain")) {
		return sniff_text_or_binary (sniffer, buffer);
	}

	if (!g_ascii_strcasecmp (content_type, "text/html"))
		return sniff_feed_or_html (sniffer, buffer);

	return g_strdup (content_type);
}

static gsize
soup_content_sniffer_real_get_buffer_size (SoupContentSniffer *sniffer)
{
	return 512;
}

static void
soup_content_sniffer_got_headers_cb (SoupMessage *msg, SoupContentSniffer *sniffer)
{
	SoupMessagePrivate *priv = SOUP_MESSAGE_GET_PRIVATE (msg);

	priv->bytes_for_sniffing = soup_content_sniffer_get_buffer_size (sniffer);
}

static void
soup_content_sniffer_request_queued (SoupSessionFeature *feature,
				     SoupSession *session,
				     SoupMessage *msg)
{
	SoupMessagePrivate *priv = SOUP_MESSAGE_GET_PRIVATE (msg);

	priv->sniffer = g_object_ref (feature);
	g_signal_connect (msg, "got-headers",
			  G_CALLBACK (soup_content_sniffer_got_headers_cb),
			  feature);
}

static void
soup_content_sniffer_request_unqueued (SoupSessionFeature *feature,
				       SoupSession *session,
				       SoupMessage *msg)
{
	SoupMessagePrivate *priv = SOUP_MESSAGE_GET_PRIVATE (msg);

	g_object_unref (priv->sniffer);
	priv->sniffer = NULL;

	g_signal_handlers_disconnect_by_func (msg, soup_content_sniffer_got_headers_cb, feature);
}

static void
soup_content_sniffer_class_init (SoupContentSnifferClass *content_sniffer_class)
{
	content_sniffer_class->sniff = soup_content_sniffer_real_sniff;
	content_sniffer_class->get_buffer_size = soup_content_sniffer_real_get_buffer_size;
}

static void
soup_content_sniffer_session_feature_init (SoupSessionFeatureInterface *feature_interface,
					   gpointer interface_data)
{
	feature_interface->request_queued = soup_content_sniffer_request_queued;
	feature_interface->request_unqueued = soup_content_sniffer_request_unqueued;
}

/**
 * soup_content_sniffer_new:
 *
 * Creates a new #SoupContentSniffer.
 *
 * Returns: a new #SoupContentSniffer
 *
 * Since: 2.28
 **/
SoupContentSniffer *
soup_content_sniffer_new ()
{
	return g_object_new (SOUP_TYPE_CONTENT_SNIFFER, NULL);
}

/**
 * soup_content_sniffer_sniff:
 * @sniffer: a #SoupContentSniffer
 * @msg: the message to sniff
 * @buffer: a buffer containing the start of @msg's response body
 * @params: (element-type utf8 utf8) (out) (transfer full) (allow-none): return
 *   location for Content-Type parameters (eg, "charset"), or %NULL
 *
 * Sniffs @buffer to determine its Content-Type. The result may also
 * be influenced by the Content-Type declared in @msg's response
 * headers.
 *
 * Return value: the sniffed Content-Type of @buffer; this will never be %NULL,
 *   but may be "application/octet-stream".
 *
 * Since: 2.28
 */
char *
soup_content_sniffer_sniff (SoupContentSniffer *sniffer,
			    SoupMessage *msg, SoupBuffer *buffer,
			    GHashTable **params)
{
	g_return_val_if_fail (SOUP_IS_CONTENT_SNIFFER (sniffer), NULL);
	g_return_val_if_fail (SOUP_IS_MESSAGE (msg), NULL);
	g_return_val_if_fail (buffer != NULL, NULL);

	return SOUP_CONTENT_SNIFFER_GET_CLASS (sniffer)->sniff (sniffer, msg, buffer, params);
}

/**
 * soup_content_sniffer_get_buffer_size:
 * @sniffer: a #SoupContentSniffer
 *
 * Gets the number of bytes @sniffer needs in order to properly sniff
 * a buffer.
 *
 * Return value: the number of bytes to sniff
 *
 * Since: 2.28
 */
gsize
soup_content_sniffer_get_buffer_size (SoupContentSniffer *sniffer)
{
	g_return_val_if_fail (SOUP_IS_CONTENT_SNIFFER (sniffer), 0);

	return SOUP_CONTENT_SNIFFER_GET_CLASS (sniffer)->get_buffer_size (sniffer);
}