Added a module to split Japanese words #3158
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
The failing tests may not be related to you code. I have the same in an unrelated change. I'm investigating. |
|
Can you please rebase your code on master? This should make the CI errors go away. |
lonvia
reviewed
Aug 20, 2023
| for p in phrases))) | ||
| return normalized | ||
|
|
||
| def split_key_japanese_phrases( |
There was a problem hiding this comment.
You probably forgot to delete the old code here.
| @@ -29,6 +29,7 @@ class BreakType(enum.Enum): | |||
| """ Break created as a result of tokenization. | |||
| This may happen in languages without spaces between words. | |||
| """ | |||
| SOFT_PHRASE = ':' | |||
There was a problem hiding this comment.
Can you please add documentation for this new type, just like it is done in the lines above.
There was a problem hiding this comment.
Thank you so much for your help and comments.
I added the documentation.
miku0
force-pushed
the
soft_phrase-final
branch
from
8f43956
to
dfbacf4
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
In these codes, Japanese addresses are divided into three categories based on administrative divisions: cities, municipalities, and below.
Nominatim uses ICU (International Components for Unicode) transliteration for user-entered addresses to split them into meaningful words. Here is an example of debugging. There are many candidates.
Fig. 1 The example of debugging.
To help make this division more accurate, when there are large administrative divisions (prefecture and city) in the string, we pre-separate them in the algorithm and put "," markers between the split words.
This "," is set to BreakType.SOFT_PHRASE in the program and words with this node are penalized with a lower search priority.
The node relationship is as follows
(1)--da->(2)--ban->(3)--shi->(4)--da->(5)--ban->(6)
|| ^^ ||
|+------大阪市--------------+ +-------大阪--------+|
+-------------------大阪市大阪---------------------+
As a result of this change, "大阪市大阪" with SOFT_PHRASE is penalized more and given lower search priority than "大阪市", the name of a city (the fifth value from the left is the penalty value).
Fig. 2 Before the change.
Fig. 3 After the change.