Added --rich-data for 'convert', and SHA1 checksum checking

10 files changed, 166 insertions, 158 deletions

diff --git a/src/convert.cc b/src/convert.cc
index 1ef3a413..da4569cc 100644
--- a/src/convert.cc
+++ b/src/convert.cc
@@ -56,72 +56,41 @@ value_t convert_command(call_scope_t& args)
   account_t * bucket  = journal.master->find_account(bucket_name);
   account_t * unknown = journal.master->find_account(_("Expenses:Unknown"));
 
-  // Make an amounts mapping for the account under consideration
-
-  typedef std::map<value_t, std::list<post_t *> > post_map_t;
-  post_map_t post_map;
-
-  xacts_iterator journal_iter(journal);
-  while (xact_t * xact = *journal_iter++) {
-    post_t * post = NULL;
-    xact_posts_iterator xact_iter(*xact);
-    while ((post = *xact_iter++) != NULL) {
-      if (post->account == bucket)
-        break;
-    }
-    if (post) {
-      post_map_t::iterator i = post_map.find(post->amount);
-      if (i == post_map.end()) {
-        std::list<post_t *> post_list;
-        post_list.push_back(post);
-        post_map.insert(post_map_t::value_type(post->amount, post_list));
-      } else {
-        (*i).second.push_back(post);
-      }
-    }
-  }
-
   // Create a flat list
   xacts_list current_xacts(journal.xacts_begin(), journal.xacts_end());
 
   // Read in the series of transactions from the CSV file
 
   print_xacts formatter(report);
-  ifstream    data(path(args.get<string>(0)));
-  csv_reader  reader(data);
-
-  while (xact_t * xact = reader.read_xact(journal, bucket)) {
-    if (report.HANDLED(invert)) {
-      foreach (post_t * post, xact->posts)
-        post->amount.in_place_negate();
-    }
+  path        csv_file_path(args.get<string>(0));
+  ifstream    data(csv_file_path);
+  csv_reader  reader(data, csv_file_path);
+
+  try {
+    while (xact_t * xact = reader.read_xact(journal, bucket,
+                                            report.HANDLED(rich_data))) {
+      if (report.HANDLED(invert)) {
+        foreach (post_t * post, xact->posts)
+          post->amount.in_place_negate();
+      }
 
-    bool matched = false;
-    if (! xact->posts.front()->amount.is_null()) {
-      post_map_t::iterator i = post_map.find(- xact->posts.front()->amount);
-      if (i != post_map.end()) {
-        std::list<post_t *>& post_list((*i).second);
-        foreach (post_t * post, post_list) {
-          if (xact->code && post->xact->code &&
-              *xact->code == *post->xact->code) {
-            matched = true;
-            break;
-          }
-          else if (xact->actual_date() == post->actual_date()) {
-            matched = true;
-            break;
-          }
-        }
+      string ref = (xact->has_tag(_("SHA1")) ?
+                    xact->get_tag(_("SHA1"))->to_string() :
+                    sha1sum(reader.get_last_line()));
+
+      checksum_map_t::const_iterator entry = journal.checksum_map.find(ref);
+      if (entry != journal.checksum_map.end()) {
+        INFO(file_context(reader.get_pathname(),
+                          reader.get_linenum())
+             << "Ignoring known SHA1 " << ref);
+        checked_delete(xact);     // ignore it
+        continue;
       }
-    }
 
-    if (matched) {
-      DEBUG("convert.csv", "Ignored xact with code: " << *xact->code);
-      checked_delete(xact);     // ignore it
-    }
-    else {
+      if (report.HANDLED(rich_data) && ! xact->has_tag(_("SHA1")))
+        xact->set_tag(_("SHA1"), string_value(ref));
+
       if (xact->posts.front()->account == NULL) {
-        // jww (2010-03-07): Bind this logic to an option: --auto-match
         if (account_t * acct =
             (report.HANDLED(auto_match) ?
              lookup_probable_account(xact->payee, current_xacts.rbegin(),
@@ -143,8 +112,16 @@ value_t convert_command(call_scope_t& args)
           formatter(*post);
       }
     }
+    formatter.flush();
+  }
+  catch (const std::exception&) {
+    add_error_context(_("While parsing file %1")
+                      << file_context(reader.get_pathname(),
+                                      reader.get_linenum()));
+    add_error_context(_("While parsing CSV line:"));
+    add_error_context(line_context(reader.get_last_line()));
+    throw;
   }
-  formatter.flush();
 
   // If not, transform the payee according to regexps
 
diff --git a/src/csv.cc b/src/csv.cc
index e2ba523d..c253f246 100644
--- a/src/csv.cc
+++ b/src/csv.cc
@@ -70,10 +70,12 @@ string csv_reader::read_field(std::istream& sin)
   else {
     while (sin.good() && ! sin.eof()) {
       sin.get(c);
-      if (c == ',')
-        break;
-      if (c != '\0')
-        field += c;
+      if (sin.good()) {
+        if (c == ',')
+          break;
+        if (c != '\0')
+          field += c;
+      }
     }
   }
   trim(field);
@@ -82,8 +84,6 @@ string csv_reader::read_field(std::istream& sin)
 
 char * csv_reader::next_line(std::istream& sin)
 {
-  static char linebuf[MAX_LINE + 1];
-
   while (sin.good() && ! sin.eof() && sin.peek() == '#')
     sin.getline(linebuf, MAX_LINE);
 
@@ -130,11 +130,13 @@ void csv_reader::read_index(std::istream& sin)
   }
 }
 
-xact_t * csv_reader::read_xact(journal_t& journal, account_t * bucket)
+xact_t * csv_reader::read_xact(journal_t& journal, account_t * bucket,
+                               bool rich_data)
 {
   char * line = next_line(in);
   if (! line || index.empty())
     return NULL;
+  linenum++;
 
   std::istringstream instr(line);
 
@@ -144,20 +146,18 @@ xact_t * csv_reader::read_xact(journal_t& journal, account_t * bucket)
   xact->set_state(item_t::CLEARED);
 
   xact->pos           = position_t();
-  xact->pos->pathname = "jww (2010-03-05): unknown";
+  xact->pos->pathname = pathname;
   xact->pos->beg_pos  = in.tellg();
-  xact->pos->beg_line = 0;
-  xact->pos->sequence = 0;
+  xact->pos->beg_line = linenum;
+  xact->pos->sequence = sequence++;
 
   post->xact = xact.get();
 
-#if 0
   post->pos           = position_t();
   post->pos->pathname = pathname;
-  post->pos->beg_pos  = line_beg_pos;
+  post->pos->beg_pos  = in.tellg();
   post->pos->beg_line = linenum;
-  post->pos->sequence = context.sequence++;
-#endif
+  post->pos->sequence = sequence++;
 
   post->set_state(item_t::CLEARED);
   post->account = NULL;
@@ -167,88 +167,80 @@ xact_t * csv_reader::read_xact(journal_t& journal, account_t * bucket)
   string total;
   string field;
 
-  try {
-    while (instr.good() && ! instr.eof()) {
-      field = read_field(instr);
+  while (instr.good() && ! instr.eof()) {
+    field = read_field(instr);
 
-      switch (index[n]) {
-      case FIELD_DATE:
-        xact->_date = parse_date(field);
-        break;
+    switch (index[n]) {
+    case FIELD_DATE:
+      xact->_date = parse_date(field);
+      break;
 
-      case FIELD_DATE_EFF:
-        xact->_date_eff = parse_date(field);
-        break;
+    case FIELD_DATE_EFF:
+      xact->_date_eff = parse_date(field);
+      break;
 
-      case FIELD_CODE:
-        if (! field.empty())
-          xact->code = field;
-        break;
+    case FIELD_CODE:
+      if (! field.empty())
+        xact->code = field;
+      break;
 
-      case FIELD_PAYEE: {
-        bool found = false;
-        foreach (payee_mapping_t& value, journal.payee_mappings) {
-          DEBUG("csv.mappings", "Looking for payee mapping: " << value.first);
-          if (value.first.match(field)) {
-            xact->payee = value.second;
-            found = true;
-            break;
-          }
+    case FIELD_PAYEE: {
+      bool found = false;
+      foreach (payee_mapping_t& value, journal.payee_mappings) {
+        DEBUG("csv.mappings", "Looking for payee mapping: " << value.first);
+        if (value.first.match(field)) {
+          xact->payee = value.second;
+          found = true;
+          break;
         }
-        if (! found)
-          xact->payee = field;
-        break;
       }
+      if (! found)
+        xact->payee = field;
+      break;
+    }
 
-      case FIELD_AMOUNT: {
-        std::istringstream amount_str(field);
-        amt.parse(amount_str, PARSE_NO_REDUCE);
-        if (! amt.has_commodity() &&
-            commodity_pool_t::current_pool->default_commodity)
-          amt.set_commodity(*commodity_pool_t::current_pool->default_commodity);
-        post->amount = amt;
-        break;
-      }
+    case FIELD_AMOUNT: {
+      std::istringstream amount_str(field);
+      amt.parse(amount_str, PARSE_NO_REDUCE);
+      if (! amt.has_commodity() &&
+          commodity_pool_t::current_pool->default_commodity)
+        amt.set_commodity(*commodity_pool_t::current_pool->default_commodity);
+      post->amount = amt;
+      break;
+    }
 
-      case FIELD_COST: {
-        std::istringstream amount_str(field);
-        amt.parse(amount_str, PARSE_NO_REDUCE);
-        if (! amt.has_commodity() &&
-            commodity_pool_t::current_pool->default_commodity)
-          amt.set_commodity
-            (*commodity_pool_t::current_pool->default_commodity);
-        post->cost = amt;
-        break;
-      }
+    case FIELD_COST: {
+      std::istringstream amount_str(field);
+      amt.parse(amount_str, PARSE_NO_REDUCE);
+      if (! amt.has_commodity() &&
+          commodity_pool_t::current_pool->default_commodity)
+        amt.set_commodity
+          (*commodity_pool_t::current_pool->default_commodity);
+      post->cost = amt;
+      break;
+    }
 
-      case FIELD_TOTAL:
-        total = field;
-        break;
+    case FIELD_TOTAL:
+      total = field;
+      break;
 
-      case FIELD_NOTE:
-        xact->note = field;
-        break;
+    case FIELD_NOTE:
+      xact->note = field;
+      break;
 
-      case FIELD_UNKNOWN:
-        if (! names[n].empty() && ! field.empty())
-          xact->set_tag(names[n], string_value(field));
-        break;
-      }
-      n++;
+    case FIELD_UNKNOWN:
+      if (! names[n].empty() && ! field.empty())
+        xact->set_tag(names[n], string_value(field));
+      break;
     }
-  }
-  catch (const std::exception&) {
-    add_error_context(_("While parsing CSV field:"));
-    add_error_context(line_context(field));
-    throw;
+    n++;
   }
 
-#if 0
-  xact->set_tag(_("Imported"),
-                string(format_date(CURRENT_DATE(), FMT_WRITTEN)));
-  xact->set_tag(_("Original"), string(line));
-  xact->set_tag(_("SHA1"), string(sha1sum(line)));
-#endif
+  if (rich_data) {
+    xact->set_tag(_("Imported"),
+                  string_value(format_date(CURRENT_DATE(), FMT_WRITTEN)));
+    xact->set_tag(_("CSV"), string_value(line));
+  }
 
   // Translate the account name, if we have enough information to do so
 
@@ -267,13 +259,11 @@ xact_t * csv_reader::read_xact(journal_t& journal, account_t * bucket)
 
   post->xact = xact.get();
 
-#if 0
   post->pos           = position_t();
   post->pos->pathname = pathname;
-  post->pos->beg_pos  = line_beg_pos;
+  post->pos->beg_pos  = in.tellg();
   post->pos->beg_line = linenum;
-  post->pos->sequence = context.sequence++;
-#endif
+  post->pos->sequence = sequence++;
 
   post->set_state(item_t::CLEARED);
   post->account = bucket;
diff --git a/src/csv.h b/src/csv.h
index 5ff8b59e..cf350e9d 100644
--- a/src/csv.h
+++ b/src/csv.h
@@ -52,9 +52,13 @@ class account_t;
 
 class csv_reader
 {
-  static const std::size_t MAX_LINE = 1024;
+  static const std::size_t MAX_LINE = 4096;
 
   std::istream& in;
+  path          pathname;
+  char          linebuf[MAX_LINE];
+  std::size_t   linenum;
+  std::size_t   sequence;
 
   enum headers_t {
     FIELD_DATE = 0,
@@ -80,13 +84,11 @@ class csv_reader
 
   std::vector<int>    index;
   std::vector<string> names;
-  std::vector<string> fields;
-
-  typedef std::map<string, string> string_map;
 
 public:
-  csv_reader(std::istream& _in)
-    : in(_in),
+  csv_reader(std::istream& _in, const path& _pathname)
+    : in(_in), pathname(_pathname),
+      linenum(0), sequence(0),
       date_mask("date"),
       date_eff_mask("posted( ?date)?"),
       code_mask("code"),
@@ -98,11 +100,30 @@ public:
     read_index(in);
   }
 
+  void   read_index(std::istream& in);
   string read_field(std::istream& in);
   char * next_line(std::istream& in);
-  void read_index(std::istream& in);
 
-  xact_t * read_xact(journal_t& journal, account_t * bucket);
+  xact_t * read_xact(journal_t& journal, account_t * bucket, bool rich_data);
+
+  const char * get_last_line() const {
+    return linebuf;
+  }
+
+  path get_pathname() const {
+    return pathname;
+  }
+  std::size_t get_linenum() const {
+    return linenum;
+  }
+
+  void reset() {
+    pathname.clear();
+    index.clear();
+    names.clear();
+    linenum = 0;
+    sequence = 0;
+  }
 };
 
 } // namespace ledger
diff --git a/src/item.cc b/src/item.cc
index 056aa04c..7184c0ef 100644
--- a/src/item.cc
+++ b/src/item.cc
@@ -72,7 +72,7 @@ bool item_t::has_tag(const mask_t& tag_mask,
   return false;
 }
 
-  optional<value_t> item_t::get_tag(const string& tag, bool) const
+optional<value_t> item_t::get_tag(const string& tag, bool) const
 {
   DEBUG("item.meta", "Getting item tag: " << tag);
   if (metadata) {
diff --git a/src/journal.cc b/src/journal.cc
index 0691954f..bbfa205c 100644
--- a/src/journal.cc
+++ b/src/journal.cc
@@ -107,6 +107,17 @@ account_t * journal_t::find_account_re(const string& regexp)
 
 bool journal_t::add_xact(xact_t * xact)
 {
+  if (optional<value_t> ref = xact->get_tag(_("SHA1"))) {
+    std::pair<checksum_map_t::iterator, bool> result
+      = checksum_map.insert(checksum_map_t::value_type(ref->to_string(), xact));
+    if (! result.second) {
+      throw_(std::runtime_error,
+             _("Found duplicated transaction with SHA1: ")
+             << ref->to_string());
+      return false;
+    }
+  }
+
   xact->journal = this;
 
   if (! xact->finalize()) {
diff --git a/src/journal.h b/src/journal.h
index ca6b6e4f..49a6292b 100644
--- a/src/journal.h
+++ b/src/journal.h
@@ -63,6 +63,7 @@ typedef std::pair<mask_t, string>      payee_mapping_t;
 typedef std::list<payee_mapping_t>     payee_mappings_t;
 typedef std::pair<mask_t, account_t *> account_mapping_t;
 typedef std::list<account_mapping_t>   account_mappings_t;
+typedef std::map<string, xact_t *>     checksum_map_t;
 
 class journal_t : public noncopyable
 {
@@ -117,6 +118,7 @@ public:
   std::list<fileinfo_t> sources;
   payee_mappings_t      payee_mappings;
   account_mappings_t    account_mappings;
+  checksum_map_t        checksum_map;
   bool                  was_loaded;
 
   journal_t();
@@ -198,6 +200,7 @@ private:
     ar & sources;
     ar & payee_mappings;
     ar & account_mappings;
+    ar & checksum_map;
   }
 #endif // HAVE_BOOST_SERIALIZATION
 };
diff --git a/src/report.cc b/src/report.cc
index b3b7233f..c562ab38 100644
--- a/src/report.cc
+++ b/src/report.cc
@@ -1094,6 +1094,7 @@ option_t<report_t> * report_t::lookup_option(const char * p)
     else OPT(revalued);
     else OPT(revalued_only);
     else OPT(revalued_total_);
+    else OPT(rich_data);
     break;
   case 's':
     OPT(sort_);
diff --git a/src/report.h b/src/report.h
index a001ffb1..565728df 100644
--- a/src/report.h
+++ b/src/report.h
@@ -313,6 +313,7 @@ public:
     HANDLER(revalued).report(out);
     HANDLER(revalued_only).report(out);
     HANDLER(revalued_total_).report(out);
+    HANDLER(rich_data).report(out);
     HANDLER(seed_).report(out);
     HANDLER(sort_).report(out);
     HANDLER(sort_all_).report(out);
@@ -893,6 +894,8 @@ public:
      set_expr(args.get<string>(0), args.get<string>(1));
    });
 
+  OPTION(report_t, rich_data);
+
   OPTION(report_t, seed_);
 
   OPTION_(report_t, sort_, DO_(args) { // -S
diff --git a/src/textual.cc b/src/textual.cc
index ddbd9943..13032236 100644
--- a/src/textual.cc
+++ b/src/textual.cc
@@ -1150,8 +1150,9 @@ post_t * instance_t::parse_post(char *          line,
 
   if (context.strict && ! post->account->has_flags(ACCOUNT_KNOWN)) {
     if (post->_state == item_t::UNCLEARED)
-      warning_(_("\"%1\", line %2: Unknown account '%3'")
-               << pathname.string() << linenum << post->account->fullname());
+      warning_(_("%1Unknown account '%2'")
+               << file_context(pathname, linenum)
+               << post->account->fullname());
     post->account->add_flags(ACCOUNT_KNOWN);
   }
 
@@ -1181,8 +1182,9 @@ post_t * instance_t::parse_post(char *          line,
       if (context.strict &&
           ! post->amount.commodity().has_flags(COMMODITY_KNOWN)) {
         if (post->_state == item_t::UNCLEARED)
-          warning_(_("\"%1\", line %2: Unknown commodity '%3'")
-                   << pathname.string() << linenum << post->amount.commodity());
+          warning_(_("%1Unknown commodity '%2'")
+                   << file_context(pathname, linenum)
+                   << post->amount.commodity());
         post->amount.commodity().add_flags(COMMODITY_KNOWN);
       }
 
diff --git a/src/utils.cc b/src/utils.cc
index 42600db3..2f64bb0a 100644
--- a/src/utils.cc
+++ b/src/utils.cc
@@ -50,8 +50,8 @@ void debug_assert(const string& reason,
                   std::size_t   line)
 {
   std::ostringstream buf;
-  buf << "Assertion failed in \"" << file << "\", line " << line
-      << ": " << func << ": " << reason;
+  buf << "Assertion failed in " << file_context(file, line)
+      << func << ": " << reason;
   throw assertion_failed(buf.str());
 }