improve handling of missing helpers messages

src/index/indexer.cpp

@@ -1,5 +1,5 @@
 #ifndef lint
-static char rcsid[] = "@(#$Id: indexer.cpp,v 1.52 2007-02-06 18:01:57 dockes Exp $ (C) 2004 J.F.Dockes";
+static char rcsid[] = "@(#$Id: indexer.cpp,v 1.53 2007-02-08 17:05:12 dockes Exp $ (C) 2004 J.F.Dockes";
 #endif
 /*
  *   This program is free software; you can redistribute it and/or modify
@@ -131,6 +131,12 @@ bool DbIndexer::indexDb(bool resetbefore, list<string> *topdirs)
 		m_dbdir.c_str()));
 	return false;
     }
+    if (!m_missingExternal.empty()) {
+	string missing;
+	stringsToString(m_missingExternal, missing);
+	LOGERR(("DbIndexer::index missing helper program(s): %s\n", 
+		missing.c_str()));
+    }
     return true;
 }
 
@@ -387,6 +393,9 @@ DbIndexer::processone(const std::string &fn, const struct stat *stp,
 	string ipath;
 	fis = interner.internfile(doc, ipath);
 	if (fis == FileInterner::FIError) {
+	    list<string> ext = interner.getMissingExternal();
+	    m_missingExternal.merge(ext);
+	    m_missingExternal.unique();
 	    // We dont stop indexing for one bad doc
 	    return FsTreeWalker::FtwOk;
 	}

src/index/indexer.h

@@ -16,7 +16,7 @@
  */
 #ifndef _INDEXER_H_INCLUDED_
 #define _INDEXER_H_INCLUDED_
-/* @(#$Id: indexer.h,v 1.22 2006-11-30 13:38:44 dockes Exp $  (C) 2004 J.F.Dockes */
+/* @(#$Id: indexer.h,v 1.23 2007-02-08 17:05:12 dockes Exp $  (C) 2004 J.F.Dockes */
 
 #include <string>
 #include <list>
@@ -143,6 +143,7 @@ class DbIndexer : public FsTreeWalkerCB {
     Rcl::Db      m_db;
     string       m_tmpdir;
     DbIxStatusUpdater *m_updater;
+    list<string> m_missingExternal; // Names of missing helper programs
 
     bool init(bool rst = false, bool rdonly = false);
 };

src/internfile/internfile.cpp

@@ -1,5 +1,5 @@
 #ifndef lint
-static char rcsid[] = "@(#$Id: internfile.cpp,v 1.26 2007-02-06 18:01:58 dockes Exp $ (C) 2004 J.F.Dockes";
+static char rcsid[] = "@(#$Id: internfile.cpp,v 1.27 2007-02-08 17:05:12 dockes Exp $ (C) 2004 J.F.Dockes";
 #endif
 /*
  *   This program is free software; you can redistribute it and/or modify
@@ -223,6 +223,37 @@ bool FileInterner::dataToTempFile(const string& dt, const string& mt,
     return true;
 }
 
+// See if the error string is formatted as a missing helper message,
+// accumulate helper name if it is
+void FileInterner::maybeExternalMissing(const string& msg)
+{
+    if (msg.find("RECFILTERROR") == 0) {
+	list<string> lerr;
+	stringToStrings(msg, lerr);
+	if (lerr.size() > 2) {
+	    list<string>::iterator it = lerr.begin();
+	    it++;
+	    if (*it == "HELPERNOTFOUND") {
+		it++;
+		m_missingExternal.push_back(it->c_str());
+	    }
+	}		    
+    }
+}
+
+const list<string>& FileInterner::getMissingExternal() 
+{
+    m_missingExternal.sort();
+    m_missingExternal.unique();
+    return m_missingExternal;
+}
+void FileInterner::getMissingExternal(string& out) 
+{
+    m_missingExternal.sort();
+    m_missingExternal.unique();
+    stringsToString(m_missingExternal, out);
+}
+
 static inline bool getKeyValue(const map<string, string>& docdata, 
 			       const string& key, string& value)
 {
@@ -368,6 +399,7 @@ FileInterner::Status FileInterner::internfile(Rcl::Doc& doc, string& ipath)
 	    Rcl::Doc doc; string ipath;
 	    collectIpathAndMT(doc, ipath);
 	    m_reason = m_handlers.back()->get_error();
+	    maybeExternalMissing(m_reason);
 	    LOGERR(("FileInterner::internfile: next_document error [%s%s%s] %s\n",
 		    m_fn.c_str(), ipath.empty()?"":"|", ipath.c_str(), 
 		    m_reason.c_str()));

src/internfile/internfile.h

@@ -16,7 +16,7 @@
  */
 #ifndef _INTERNFILE_H_INCLUDED_
 #define _INTERNFILE_H_INCLUDED_
-/* @(#$Id: internfile.h,v 1.13 2007-02-08 12:25:49 dockes Exp $  (C) 2004 J.F.Dockes */
+/* @(#$Id: internfile.h,v 1.14 2007-02-08 17:05:12 dockes Exp $  (C) 2004 J.F.Dockes */
 
 #include <string>
 #include <vector>
@@ -91,6 +91,8 @@ class FileInterner {
 			     const string& ipath, const string& mtype);
 
     const string& getReason() const {return m_reason;}
+    const list<string>& getMissingExternal();
+    void getMissingExternal(string& missing);
 
  private:
     static const unsigned int MAXHANDLERS = 20;
@@ -110,12 +112,15 @@ class FileInterner {
     vector<TempFile>       m_tempfiles;
     // Error data if any
     string                 m_reason;
+    // Missing external programs
+    list<string>           m_missingExternal;
 
     void tmpcleanup();
     bool dijontorcl(Rcl::Doc&);
     void collectIpathAndMT(Rcl::Doc&, string& ipath) const;
     bool dataToTempFile(const string& data, const string& mt, string& fn);
     void popHandler();
+    void maybeExternalMissing(const string& msg);
 };
 
  

src/qtgui/preview_w.cpp

@@ -1,5 +1,5 @@
 #ifndef lint
-static char rcsid[] = "@(#$Id: preview_w.cpp,v 1.15 2007-02-06 18:01:58 dockes Exp $ (C) 2005 J.F.Dockes";
+static char rcsid[] = "@(#$Id: preview_w.cpp,v 1.16 2007-02-08 17:05:12 dockes Exp $ (C) 2005 J.F.Dockes";
 #endif
 /*
  *   This program is free software; you can redistribute it and/or modify
@@ -443,7 +443,7 @@ class LoadThread : public QThread {
     string tmpdir;
     int loglevel;
  public: 
-    string reason;
+    string missing;
     LoadThread(int *stp, Rcl::Doc *odoc, string fn, string ip, string *mt) 
 	: statusp(stp), out(odoc), filename(fn), ipath(ip), mtype(mt) 
 	{
@@ -457,6 +457,7 @@ class LoadThread : public QThread {
     }
     virtual void run() {
 	DebugLog::getdbl()->setloglevel(loglevel);
+	string reason;
 	if (!maketmpdir(tmpdir, reason)) {
 	    QMessageBox::critical(0, "Recoll",
 				  Preview::tr("Cannot create temporary directory"));
@@ -486,7 +487,7 @@ class LoadThread : public QThread {
 		*statusp = 0;
 	    } else {
 		out->mimetype = interner.getMimetype();
-		reason = interner.getReason();
+		interner.getMissingExternal(missing);
 		*statusp = -1;
 	    }
 	} catch (CancelExcept) {
@@ -588,19 +589,10 @@ bool Preview::loadFileInCurrentTab(string fn, size_t sz, const Rcl::Doc &idoc,
 	return false;
     if (status != 0) {
 	QString explain;
-	if (lthr.reason.find("RECFILTERROR") == 0) {
+	if (!lthr.missing.empty()) {
-	    list<string> lerr;
+	    explain = QString::fromAscii("<br>") +
-	    stringToStrings(lthr.reason, lerr);
+		tr("Missing helper program: ") + 
-	    if (lerr.size() > 2) {
+		QString::fromLocal8Bit(lthr.missing.c_str());
-		list<string>::iterator it = lerr.begin();
-		it++;
-		if (*it == "HELPERNOTFOUND") {
-		    it++;
-		    explain = QString::fromAscii("<br>") +
-			tr("Missing helper program: ") + 
-			QString::fromLocal8Bit(it->c_str());
-		}
-	    }		    
 	}
 	QMessageBox::warning(0, "Recoll",
 			     tr("Can't turn doc into internal "

src/utils/smallut.cpp

@@ -1,5 +1,5 @@
 #ifndef lint
-static char rcsid[] = "@(#$Id: smallut.cpp,v 1.25 2006-12-19 12:11:21 dockes Exp $ (C) 2004 J.F.Dockes";
+static char rcsid[] = "@(#$Id: smallut.cpp,v 1.26 2007-02-08 17:05:12 dockes Exp $ (C) 2004 J.F.Dockes";
 #endif
 /*
  *   This program is free software; you can redistribute it and/or modify
@@ -258,6 +258,31 @@ bool stringToStrings(const string &s, std::list<string> &tokens)
     return true;
 }
 
+void stringsToString(const list<string> &tokens, string &s) 
+{
+    for (list<string>::const_iterator it = tokens.begin();
+	 it != tokens.end(); it++) {
+	bool hasblanks = false;
+	if (it->find_first_of(" \t\n") != string::npos)
+	    hasblanks = true;
+	if (it != tokens.begin())
+	    s.append(1, ' ');
+	if (hasblanks)
+	    s.append(1, '"');
+	for (unsigned int i = 0; i < it->length(); i++) {
+	    char car = it->at(i);
+	    if (car == '"') {
+		s.append(1, '\\');
+		s.append(1, car);
+	    } else {
+		s.append(1, car);
+	    }
+	}
+	if (hasblanks)
+	    s.append(1, '"');
+    }
+}
+
 void stringToTokens(const string& str, list<string>& tokens,
 		    const string& delims, bool skipinit)
 {

src/utils/smallut.h

@@ -16,7 +16,7 @@
  */
 #ifndef _SMALLUT_H_INCLUDED_
 #define _SMALLUT_H_INCLUDED_
-/* @(#$Id: smallut.h,v 1.25 2006-12-19 12:11:21 dockes Exp $  (C) 2004 J.F.Dockes */
+/* @(#$Id: smallut.h,v 1.26 2007-02-08 17:05:12 dockes Exp $  (C) 2004 J.F.Dockes */
 #include <string>
 #include <list>
 #include <map>
@@ -45,6 +45,10 @@ extern bool samecharset(const string &cs1, const string &cs2);
  * dquotes can be escaped with \ etc...
  */
 extern bool stringToStrings(const string &s, list<string> &tokens);
+/**
+ * Inverse operation:
+ */
+extern void stringsToString(const list<string> &tokens, string &s);
 
 /**
  * Split input string. No handling of quoting