Merge pull request #213 from NationalLibraryOfNorway/feat/add-fields-…

…corpusbuilder add two fields (publisher,city) to the corpus builder
NationalLibraryOfNorway · Aug 2, 2024 · 5c5da3c · 5c5da3c
2 parents 031bfe0 + 86a828d
commit 5c5da3c
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 0 deletions.
diff --git a/dhlab/api/dhlab_api.py b/dhlab/api/dhlab_api.py
@@ -637,6 +637,8 @@ def document_corpus(
     title: str = None,
     ddk: str = None,
     subject: str = None,
+    publisher: str = None,
+    city: str = None,
     lang: str = None,
     limit: int = None,
     order_by: str = None,
@@ -660,6 +662,8 @@ def document_corpus(
     :param str ddk: `Dewey Decimal Classification
         <https://no.wikipedia.org/wiki/Deweys_desimalklassifikasjon>`_ identifier.
     :param str subject: subject (keywords) of the publication.
+    :param str publisher: Name of publisher.
+    :param str city: place of publication.
     :param str lang: Language of the publication, as a 3-letter ISO code.
         Example: ``"nob"`` or ``"nno"``
     :param int limit: number of items to sample.

diff --git a/dhlab/text/corpus.py b/dhlab/text/corpus.py
@@ -41,6 +41,8 @@ def __init__(
         title=None,
         ddk=None,
         subject=None,
+        publisher=None,
+        city=None,
         lang=None,
         limit=10,
         order_by="random",
@@ -65,6 +67,8 @@ def __init__(
             <https://no.wikipedia.org/wiki/Deweys_desimalklassifikasjon>`\
                 _ identifier.
         :param str subject: subject (keywords) of the publication.
+        :param str publisher: Name of publisher.
+        :param str city: place of publication.
         :param str lang: Language of the publication, as a 3-letter ISO code.
             Example: ``"nob"`` or ``"nno"``
         :param int limit: number of items to sample.
@@ -82,6 +86,8 @@ def __init__(
             or title
             or ddk
             or lang
+            or publisher
+            or city
         ):
             self.corpus = document_corpus(
                 doctype,
@@ -95,6 +101,8 @@ def __init__(
                 title,
                 ddk,
                 subject,
+                publisher,
+                city,
                 lang,
                 limit,
                 order_by,