Merge remote-tracking branch 'upstream/master'

[nominatim.git] / nominatim / tokenizer / icu_rule_loader.py

diff --git a/nominatim/tokenizer/icu_rule_loader.py b/nominatim/tokenizer/icu_rule_loader.py
index 269faed981abbbb9ffc530bd32d6b38ae0c30df4..b8551038aa42283dfccfc6bedaf1c0b89c0ba68b 100644 (file)
--- a/nominatim/tokenizer/icu_rule_loader.py
+++ b/nominatim/tokenizer/icu_rule_loader.py
@@ -1,167 +1,160 @@
 """
 Helper class to create ICU rules from a configuration file.
 """
 """
 Helper class to create ICU rules from a configuration file.
 """

+import importlib

 import io
 import io

+import json

 import logging
 import logging

-from collections import defaultdict
-import itertools
-
-import yaml
-from icu import Transliterator

 
 

+from nominatim.config import flatten_config_list
+from nominatim.db.properties import set_property, get_property

 from nominatim.errors import UsageError
 from nominatim.errors import UsageError

+from nominatim.tokenizer.place_sanitizer import PlaceSanitizer
+from nominatim.tokenizer.icu_token_analysis import ICUTokenAnalysis
+import nominatim.tools.country_info

 
 LOG = logging.getLogger()
 
 
 LOG = logging.getLogger()
 

+DBCFG_IMPORT_NORM_RULES = "tokenizer_import_normalisation"
+DBCFG_IMPORT_TRANS_RULES = "tokenizer_import_transliteration"
+DBCFG_IMPORT_ANALYSIS_RULES = "tokenizer_import_analysis_rules"
+
+
+def _get_section(rules, section):
+    """ Get the section named 'section' from the rules. If the section does
+        not exist, raise a usage error with a meaningful message.
+    """
+    if section not in rules:
+        LOG.fatal("Section '%s' not found in tokenizer config.", section)
+        raise UsageError("Syntax error in tokenizer configuration file.")
+
+    return rules[section]
+

 
 class ICURuleLoader:
     """ Compiler for ICU rules from a tokenizer configuration file.
     """
 
 
 class ICURuleLoader:
     """ Compiler for ICU rules from a tokenizer configuration file.
     """
 

-    def __init__(self, configfile):
-        self.configfile = configfile
-        self.compound_suffixes = set()
-        self.abbreviations = defaultdict()
+    def __init__(self, config):
+        rules = config.load_sub_configuration('icu_tokenizer.yaml',
+                                              config='TOKENIZER_CONFIG')
+
+        # Make sure country information is available to analyzers and sanatizers.
+        nominatim.tools.country_info.setup_country_config(config)
+
+        self.normalization_rules = self._cfg_to_icu_rules(rules, 'normalization')
+        self.transliteration_rules = self._cfg_to_icu_rules(rules, 'transliteration')
+        self.analysis_rules = _get_section(rules, 'token-analysis')
+        self._setup_analysis()
+
+        # Load optional sanitizer rule set.
+        self.sanitizer_rules = rules.get('sanitizers', [])
+

 
 

-        if configfile.suffix == '.yaml':
-            self._load_from_yaml()
-        else:
-            raise UsageError("Unknown format of tokenizer configuration.")
+    def load_config_from_db(self, conn):
+        """ Get previously saved parts of the configuration from the
+            database.
+        """
+        self.normalization_rules = get_property(conn, DBCFG_IMPORT_NORM_RULES)
+        self.transliteration_rules = get_property(conn, DBCFG_IMPORT_TRANS_RULES)
+        self.analysis_rules = json.loads(get_property(conn, DBCFG_IMPORT_ANALYSIS_RULES))
+        self._setup_analysis()
+
+
+    def save_config_to_db(self, conn):
+        """ Save the part of the configuration that cannot be changed into
+            the database.
+        """
+        set_property(conn, DBCFG_IMPORT_NORM_RULES, self.normalization_rules)
+        set_property(conn, DBCFG_IMPORT_TRANS_RULES, self.transliteration_rules)
+        set_property(conn, DBCFG_IMPORT_ANALYSIS_RULES, json.dumps(self.analysis_rules))
+
+
+    def make_sanitizer(self):
+        """ Create a place sanitizer from the configured rules.
+        """
+        return PlaceSanitizer(self.sanitizer_rules)
+
+
+    def make_token_analysis(self):
+        """ Create a token analyser from the reviouly loaded rules.
+        """
+        return ICUTokenAnalysis(self.normalization_rules,
+                                self.transliteration_rules, self.analysis)

 
 
     def get_search_rules(self):
         """ Return the ICU rules to be used during search.
 
 
     def get_search_rules(self):
         """ Return the ICU rules to be used during search.

-            The rules combine normalization, compound decomposition (including
-            abbreviated compounds) and transliteration.
+            The rules combine normalization and transliteration.

         """
         # First apply the normalization rules.
         rules = io.StringIO()
         rules.write(self.normalization_rules)
 
         """
         # First apply the normalization rules.
         rules = io.StringIO()
         rules.write(self.normalization_rules)
 

-        # For all compound suffixes: add them in their full and any abbreviated form.
-        suffixes = set()
-        for suffix in self.compound_suffixes:
-            suffixes.add(suffix)
-            suffixes.update(self.abbreviations.get(suffix, []))
-
-        for suffix in sorted(suffixes, key=len, reverse=True):
-            rules.write("'{0} ' > ' {0} ';".format(suffix))
-
-        # Finally add transliteration.
+        # Then add transliteration.

         rules.write(self.transliteration_rules)
         return rules.getvalue()
 
         rules.write(self.transliteration_rules)
         return rules.getvalue()
 

+

     def get_normalization_rules(self):
         """ Return rules for normalisation of a term.
         """
         return self.normalization_rules
 
     def get_normalization_rules(self):
         """ Return rules for normalisation of a term.
         """
         return self.normalization_rules
 

+

     def get_transliteration_rules(self):
         """ Return the rules for converting a string into its asciii representation.
         """
         return self.transliteration_rules
 
     def get_transliteration_rules(self):
         """ Return the rules for converting a string into its asciii representation.
         """
         return self.transliteration_rules
 

-    def get_replacement_pairs(self):
-        """ Return the list of possible compound decompositions with
-            application of abbreviations included.
-            The result is a list of pairs: the first item is the sequence to
-            replace, the second is a list of replacements.
-        """
-        synonyms = defaultdict(set)
-
-        for full, abbr in self.abbreviations.items():
-            key = ' ' + full + ' '
-            # Entries in the abbreviation list always apply to full words:
-            synonyms[key].update((' ' + a + ' ' for a in abbr))
-            # Replacements are optional, so add a noop
-            synonyms[key].add(key)
-
-        # Entries in the compound list expand to themselves and to
-        # abbreviations.
-        for suffix in self.compound_suffixes:
-            keyset = synonyms[suffix + ' ']
-            keyset.add(' ' + suffix + ' ')
-            keyset.update((' ' + a + ' ' for a in self.abbreviations.get(suffix, [])))
-            # The terms the entries are shortended to, need to be decompunded as well.
-            for abbr in self.abbreviations.get(suffix, []):
-                synonyms[abbr + ' '].add(' ' + abbr + ' ')
-
-        # sort the resulting list by descending length (longer matches are prefered).
-        sorted_keys = sorted(synonyms.keys(), key=len, reverse=True)
-
-        return [(k, list(synonyms[k])) for k in sorted_keys]
-
-
-    def _load_from_yaml(self):
-        rules = yaml.safe_load(self.configfile.read_text())
-
-        self.normalization_rules = self._cfg_to_icu_rules(rules, 'normalization')
-        self.transliteration_rules = self._cfg_to_icu_rules(rules, 'transliteration')
-        self._parse_compound_suffix_list(self._get_section(rules, 'compound_suffixes'))
-        self._parse_abbreviation_list(self._get_section(rules, 'abbreviations'))

 
 

-
-    def _get_section(self, rules, section):
-        """ Get the section named 'section' from the rules. If the section does
-            not exist, raise a usage error with a meaningful message.
+    def _setup_analysis(self):
+        """ Process the rules used for creating the various token analyzers.

         """
         """

-        if section not in rules:
-            LOG.fatal("Section '%s' not found in tokenizer config '%s'.",
-                      section, str(self.configfile))
-            raise UsageError("Syntax error in tokenizer configuration file.")
+        self.analysis = {}
+
+        if not isinstance(self.analysis_rules, list):
+            raise UsageError("Configuration section 'token-analysis' must be a list.")

 
 

-        return rules[section]
+        for section in self.analysis_rules:
+            name = section.get('id', None)
+            if name in self.analysis:
+                if name is None:
+                    LOG.fatal("ICU tokenizer configuration has two default token analyzers.")
+                else:
+                    LOG.fatal("ICU tokenizer configuration has two token "
+                              "analyzers with id '%s'.", name)
+                raise UsageError("Syntax error in ICU tokenizer config.")
+            self.analysis[name] = TokenAnalyzerRule(section, self.normalization_rules)

 
 
 
 

-    def _cfg_to_icu_rules(self, rules, section):
+    @staticmethod
+    def _cfg_to_icu_rules(rules, section):

         """ Load an ICU ruleset from the given section. If the section is a
             simple string, it is interpreted as a file name and the rules are
             loaded verbatim from the given file. The filename is expected to be
             relative to the tokenizer rule file. If the section is a list then
             each line is assumed to be a rule. All rules are concatenated and returned.
         """
         """ Load an ICU ruleset from the given section. If the section is a
             simple string, it is interpreted as a file name and the rules are
             loaded verbatim from the given file. The filename is expected to be
             relative to the tokenizer rule file. If the section is a list then
             each line is assumed to be a rule. All rules are concatenated and returned.
         """

-        content = self._get_section(rules, section)
+        content = _get_section(rules, section)

 
         if content is None:
             return ''
 
 
         if content is None:
             return ''
 

-        if isinstance(content, str):
-            return (self.configfile.parent / content).read_text().replace('\n', ' ')
-
-        return ';'.join(content) + ';'
-
-
-    def _parse_compound_suffix_list(self, rules):
-        if not rules:
-            self.compound_suffixes = set()
-            return
-
-        norm = Transliterator.createFromRules("rule_loader_normalization",
-                                              self.normalization_rules)
+        return ';'.join(flatten_config_list(content, section)) + ';'

 
 

-        # Make sure all suffixes are in their normalised form.
-        self.compound_suffixes = set((norm.transliterate(s) for s in rules))

 
 

+class TokenAnalyzerRule:
+    """ Factory for a single analysis module. The class saves the configuration
+        and creates a new token analyzer on request.
+    """

 
 

-    def _parse_abbreviation_list(self, rules):
-        self.abbreviations = defaultdict(list)
-
-        if not rules:
-            return
-
-        norm = Transliterator.createFromRules("rule_loader_normalization",
-                                              self.normalization_rules)
-
-        for rule in rules:
-            parts = rule.split('=>')
-            if len(parts) != 2:
-                LOG.fatal("Syntax error in abbreviation section, line: %s", rule)
-                raise UsageError("Syntax error in tokenizer configuration file.")
-
-            # Make sure all terms match the normalised version.
-            fullterms = (norm.transliterate(t.strip()) for t in parts[0].split(','))
-            abbrterms = (norm.transliterate(t.strip()) for t in parts[1].split(','))
+    def __init__(self, rules, normalization_rules):
+        # Find the analysis module
+        module_name = 'nominatim.tokenizer.token_analysis.' \
+                      + _get_section(rules, 'analyzer').replace('-', '_')
+        analysis_mod = importlib.import_module(module_name)
+        self.create = analysis_mod.create

 
 

-            for full, abbr in itertools.product(fullterms, abbrterms):
-                if full and abbr:
-                    self.abbreviations[full].append(abbr)
+        # Load the configuration.
+        self.config = analysis_mod.configure(rules, normalization_rules)