import dataclasses
import enum
-class BreakType(enum.Enum):
- """ Type of break between tokens.
- """
- START = '<'
- """ Begin of the query. """
- END = '>'
- """ End of the query. """
- PHRASE = ','
- """ Break between two phrases. """
- WORD = ' '
- """ Break between words. """
- PART = '-'
- """ Break inside a word, for example a hyphen or apostrophe. """
- TOKEN = '`'
- """ Break created as a result of tokenization.
- This may happen in languages without spaces between words.
- """
+BreakType = str
+""" Type of break between tokens.
+"""
+BREAK_START = '<'
+""" Begin of the query. """
+BREAK_END = '>'
+""" End of the query. """
+BREAK_PHRASE = ','
+""" Hard break between two phrases. Address parts cannot cross hard
+ phrase boundaries."""
+BREAK_SOFT_PHRASE = ':'
+""" Likely break between two phrases. Address parts should not cross soft
+ phrase boundaries. Soft breaks can be inserted by a preprocessor
+ that is analysing the input string.
+"""
+BREAK_WORD = ' '
+""" Break between words. """
+BREAK_PART = '-'
+""" Break inside a word, for example a hyphen or apostrophe. """
+BREAK_TOKEN = '`'
+""" Break created as a result of tokenization.
+ This may happen in languages without spaces between words.
+"""
-class TokenType(enum.Enum):
- """ Type of token.
- """
- WORD = enum.auto()
- """ Full name of a place. """
- PARTIAL = enum.auto()
- """ Word term without breaks, does not necessarily represent a full name. """
- HOUSENUMBER = enum.auto()
- """ Housenumber term. """
- POSTCODE = enum.auto()
- """ Postal code term. """
- COUNTRY = enum.auto()
- """ Country name or reference. """
- QUALIFIER = enum.auto()
- """ Special term used together with name (e.g. _Hotel_ Bellevue). """
- NEAR_ITEM = enum.auto()
- """ Special term used as searchable object(e.g. supermarket in ...). """
+
+TokenType = str
+""" Type of token.
+"""
+TOKEN_WORD = 'W'
+""" Full name of a place. """
+TOKEN_PARTIAL = 'w'
+""" Word term without breaks, does not necessarily represent a full name. """
+TOKEN_HOUSENUMBER = 'H'
+""" Housenumber term. """
+TOKEN_POSTCODE = 'P'
+""" Postal code term. """
+TOKEN_COUNTRY = 'C'
+""" Country name or reference. """
+TOKEN_QUALIFIER = 'Q'
+""" Special term used together with name (e.g. _Hotel_ Bellevue). """
+TOKEN_NEAR_ITEM = 'N'
+""" Special term used as searchable object(e.g. supermarket in ...). """
class PhraseType(enum.Enum):
""" Check if the given token type can be used with the phrase type.
"""
if self == PhraseType.NONE:
- return not is_full_phrase or ttype != TokenType.QUALIFIER
+ return not is_full_phrase or ttype != TOKEN_QUALIFIER
if self == PhraseType.AMENITY:
- return ttype in (TokenType.WORD, TokenType.PARTIAL)\
- or (is_full_phrase and ttype == TokenType.NEAR_ITEM)\
- or (not is_full_phrase and ttype == TokenType.QUALIFIER)
+ return ttype in (TOKEN_WORD, TOKEN_PARTIAL)\
+ or (is_full_phrase and ttype == TOKEN_NEAR_ITEM)\
+ or (not is_full_phrase and ttype == TOKEN_QUALIFIER)
if self == PhraseType.STREET:
- return ttype in (TokenType.WORD, TokenType.PARTIAL, TokenType.HOUSENUMBER)
+ return ttype in (TOKEN_WORD, TOKEN_PARTIAL, TOKEN_HOUSENUMBER)
if self == PhraseType.POSTCODE:
- return ttype == TokenType.POSTCODE
+ return ttype == TOKEN_POSTCODE
if self == PhraseType.COUNTRY:
- return ttype == TokenType.COUNTRY
+ return ttype == TOKEN_COUNTRY
- return ttype in (TokenType.WORD, TokenType.PARTIAL)
+ return ttype in (TOKEN_WORD, TOKEN_PARTIAL)
@dataclasses.dataclass
count: int
addr_count: int
lookup_word: str
- is_indexed: bool
-
@abstractmethod
def get_category(self) -> Tuple[str, str]:
category objects.
"""
+
@dataclasses.dataclass
class TokenRange:
""" Indexes of query nodes over which a token spans.
"""
start: int
end: int
+ penalty: Optional[float] = None
def __lt__(self, other: 'TokenRange') -> bool:
return self.end <= other.start
-
def __le__(self, other: 'TokenRange') -> bool:
return NotImplemented
-
def __gt__(self, other: 'TokenRange') -> bool:
return self.start >= other.end
-
def __ge__(self, other: 'TokenRange') -> bool:
return NotImplemented
-
def replace_start(self, new_start: int) -> 'TokenRange':
""" Return a new token range with the new start.
"""
return TokenRange(new_start, self.end)
-
def replace_end(self, new_end: int) -> 'TokenRange':
""" Return a new token range with the new end.
"""
return TokenRange(self.start, new_end)
-
def split(self, index: int) -> Tuple['TokenRange', 'TokenRange']:
""" Split the span into two spans at the given index.
The index must be within the span.
ttype: TokenType
tokens: List[Token]
-
def add_penalty(self, penalty: float) -> None:
""" Add the given penalty to all tokens in the list.
"""
"""
return any(tl.end == end and tl.ttype in ttypes for tl in self.starting)
-
def get_tokens(self, end: int, ttype: TokenType) -> Optional[List[Token]]:
""" Get the list of tokens of the given type starting at this node
and ending at the node 'end'. Returns 'None' if no such
def __init__(self, source: List[Phrase]) -> None:
self.source = source
self.nodes: List[QueryNode] = \
- [QueryNode(BreakType.START, source[0].ptype if source else PhraseType.NONE)]
-
+ [QueryNode(BREAK_START, source[0].ptype if source else PhraseType.NONE)]
def num_token_slots(self) -> int:
""" Return the length of the query in vertice steps.
"""
return len(self.nodes) - 1
-
def add_node(self, btype: BreakType, ptype: PhraseType) -> None:
""" Append a new break node with the given break type.
The phrase type denotes the type for any tokens starting
"""
self.nodes.append(QueryNode(btype, ptype))
-
def add_token(self, trange: TokenRange, ttype: TokenType, token: Token) -> None:
""" Add a token to the query. 'start' and 'end' are the indexes of the
nodes from which to which the token spans. The indexes must exist
be added to, then the token is silently dropped.
"""
snode = self.nodes[trange.start]
- full_phrase = snode.btype in (BreakType.START, BreakType.PHRASE)\
- and self.nodes[trange.end].btype in (BreakType.PHRASE, BreakType.END)
+ full_phrase = snode.btype in (BREAK_START, BREAK_PHRASE)\
+ and self.nodes[trange.end].btype in (BREAK_PHRASE, BREAK_END)
if snode.ptype.compatible_with(ttype, full_phrase):
tlist = snode.get_tokens(trange.end, ttype)
if tlist is None:
else:
tlist.append(token)
-
def get_tokens(self, trange: TokenRange, ttype: TokenType) -> List[Token]:
""" Get the list of tokens of a given type, spanning the given
nodes. The nodes must exist. If no tokens exist, an
"""
return self.nodes[trange.start].get_tokens(trange.end, ttype) or []
-
def get_partials_list(self, trange: TokenRange) -> List[Token]:
""" Create a list of partial tokens between the given nodes.
The list is composed of the first token of type PARTIAL
going to the subsequent node. Such PARTIAL tokens are
assumed to exist.
"""
- return [next(iter(self.get_tokens(TokenRange(i, i+1), TokenType.PARTIAL)))
- for i in range(trange.start, trange.end)]
-
+ return [next(iter(self.get_tokens(TokenRange(i, i+1), TOKEN_PARTIAL)))
+ for i in range(trange.start, trange.end)]
def iter_token_lists(self) -> Iterator[Tuple[int, QueryNode, TokenList]]:
""" Iterator over all token lists in the query.
for tlist in node.starting:
yield i, node, tlist
-
def find_lookup_word_by_id(self, token: int) -> str:
""" Find the first token with the given token ID and return
its lookup word. Returns 'None' if no such token exists.
for tlist in node.starting:
for t in tlist.tokens:
if t.token == token:
- return f"[{tlist.ttype.name[0]}]{t.lookup_word}"
+ return f"[{tlist.ttype}]{t.lookup_word}"
return 'None'
projects / nominatim.git / blobdiff