recoll/src/internfile/mh_html.cpp

#ifndef lint
static char rcsid[] = "@(#$Id: mh_html.cpp,v 1.26 2008-10-03 06:17:46 dockes Exp $ (C) 2005 J.F.Dockes";
#endif
/*
 *   This program is free software; you can redistribute it and/or modify
 *   it under the terms of the GNU General Public License as published by
 *   the Free Software Foundation; either version 2 of the License, or
 *   (at your option) any later version.
 *
 *   This program is distributed in the hope that it will be useful,
 *   but WITHOUT ANY WARRANTY; without even the implied warranty of
 *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *   GNU General Public License for more details.
 *
 *   You should have received a copy of the GNU General Public License
 *   along with this program; if not, write to the
 *   Free Software Foundation, Inc.,
 *   59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
 */


#include "mimehandler.h"
#include "debuglog.h"
#include "csguess.h"
#include "readfile.h"
#include "transcode.h"
#include "mimeparse.h"
#include "myhtmlparse.h"
#include "indextext.h"
#include "mh_html.h"
#include "smallut.h"
#include "md5.h"

#include <iostream>

#ifndef NO_NAMESPACES
using namespace std;
#endif /* NO_NAMESPACES */


bool MimeHandlerHtml::set_document_file(const string &fn)
{
    LOGDEB0(("textHtmlToDoc: %s\n", fn.c_str()));
    RecollFilter::set_document_file(fn);
    string otext;
    if (!file_to_string(fn, otext)) {
	LOGINFO(("textHtmlToDoc: cant read: %s\n", fn.c_str()));
	return false;
    }
    m_filename = fn;
    return set_document_string(otext);
}

bool MimeHandlerHtml::set_document_string(const string& htext) 
{
    m_html = htext;
    m_havedoc = true;

    // We want to compute the md5 now because we may modify m_html later
    string md5, xmd5;
    MD5String(htext, md5);
    m_metaData["md5"] = MD5HexPrint(md5, xmd5);

    return true;
}

bool MimeHandlerHtml::next_document()
{
    if (m_havedoc == false)
	return false;
    m_havedoc = false;
    // If set_doc(fn), take note of file name.
    string fn = m_filename;
    m_filename.erase();

    string charset = m_defcharset;
    LOGDEB(("textHtmlToDoc: next_document. defcharset before parsing: [%s]\n", 
	    charset.c_str()));

    // - We first try to convert from the default configured charset
    //   (which may depend of the current directory) to utf-8. If this
    //   fails, we keep the original text
    // - During parsing, if we find a charset parameter, and it differs from
    //   what we started with, we abort and restart with the parameter value
    //   instead of the configuration one.

    MyHtmlParser result;
    for (int pass = 0; pass < 2; pass++) {
	string transcoded;
	LOGDEB(("Html::mkDoc: pass %d\n", pass));
	MyHtmlParser p;

	// Try transcoding. If it fails, use original text.
	int ecnt;
	if (!transcode(m_html, transcoded, charset, "UTF-8", &ecnt)) {
	    LOGDEB(("textHtmlToDoc: transcode failed from cs '%s' to UTF-8 for"
		    "[%s]", charset.c_str(), fn.empty()?"unknown":fn.c_str()));
	    transcoded = m_html;
	    // We don't know the charset, at all
	    p.reset_charsets();
	    charset = "";
	} else {
	    if (ecnt) {
		if (pass == 0) {
		    LOGDEB(("textHtmlToDoc: init transcode had %d errors for "
			    "[%s]\n", ecnt, fn.empty()?"unknown":fn.c_str()));
		} else {
		    LOGERR(("textHtmlToDoc: final transcode had %d errors for "
			    "[%s]\n", ecnt, fn.empty()?"unknown":fn.c_str()));
		}
	    }
	    // charset has the putative source charset, transcoded is now
	    // in utf-8
	    p.set_charsets(charset, "utf-8");
	}

	try {
	    p.parse_html(transcoded);
	    // No exception: ok?
	    result = p;
	    break;
	} catch (bool diag) {
	    result = p;
	    if (diag == true) {
		// Parser throws true at end of text. ok

		if (m_forPreview) {
		    // Save the html text
		    m_html = transcoded;
		    // In many cases, we need to change the charset decl,
		    // because the file was transcoded. It seems that just
		    // inserting one is enough (only the 1st one seems to
		    // be used by browsers/qtextedit).
                    string::size_type idx = m_html.find("<head>");
		    if (idx == string::npos)
			idx = m_html.find("<HEAD>");
		    if (idx != string::npos)
			m_html.replace(idx+6, 0, 
				       "<meta http-equiv=\"content-type\" "
				       "content=\"text/html; charset=utf-8\">");
		}

		break;
	    }

	    LOGDEB(("textHtmlToDoc: charset [%s] doc charset [%s]\n",
		    charset.c_str(), result.get_charset().c_str()));
	    if (!result.get_charset().empty() && 
		!samecharset(result.get_charset(), result.fromcharset)) {
		LOGDEB(("textHtmlToDoc: reparse for charsets\n"));
		// Set the origin charset as specified in document before
		// transcoding again
		charset = result.get_charset();
	    } else {
		LOGERR(("textHtmlToDoc:: error: non charset exception\n"));
		return false;
	    }
	}
    }

    m_metaData["origcharset"] = result.get_charset();
    m_metaData["content"] = result.dump;
    m_metaData["charset"] = "utf-8";
    // Avoid setting empty values which would crush ones possibly inherited
    // from parent (if we're an attachment)
    if (!result.dmtime.empty())
	m_metaData["modificationdate"] = result.dmtime;
    m_metaData["mimetype"] = "text/plain";

    for (map<string,string>::const_iterator it = result.meta.begin(); 
	 it != result.meta.end(); it++) {
	if (!it->second.empty())
	    m_metaData[it->first] = it->second;
    }
    return true;
}