Name the pluck CSV file according to the passed arguments

.gitignore

@@ -9,5 +9,5 @@ venv/
 /data
 /docs/_build
 /htmlcov
-/pluck.csv
+/pluck-*.csv
 /pluck_skipped.json

kingfisher_scrapy/commands/pluck.py

@@ -8,6 +8,8 @@
 from scrapy.crawler import CrawlerProcess
 from scrapy.exceptions import UsageError
 
+from kingfisher_scrapy.util import _pluck_filename
+
 logger = logging.getLogger(__name__)
 
 
@@ -33,8 +35,9 @@ def run(self, args, opts):
         # Limit concurrent requests, to download the minimum.
         self.settings.set('CONCURRENT_REQUESTS', 1)
 
-        if os.path.isfile('pluck.csv'):
-            os.unlink('pluck.csv')
+        filename = _pluck_filename(opts)
+        if os.path.isfile(filename):
+            os.unlink(filename)
 
         runner = CrawlerProcess(settings=self.settings)
 

kingfisher_scrapy/extensions.py

@@ -9,20 +9,20 @@
 
 from kingfisher_scrapy.items import File, FileError, FileItem, PluckedItem
 from kingfisher_scrapy.kingfisher_process import Client
+from kingfisher_scrapy.util import _pluck_filename
 
 
 # https://docs.scrapy.org/en/latest/topics/extensions.html#writing-your-own-extension
 class KingfisherPluck:
-    def __init__(self, filename):
-        self.filename = filename
+    def __init__(self, directory):
+        self.directory = directory
         self.spiders_seen = set()
 
     @classmethod
     def from_crawler(cls, crawler):
-        path = crawler.settings['KINGFISHER_PLUCK_PATH']
-        filename = os.path.join(path, 'pluck.csv')
+        directory = crawler.settings['KINGFISHER_PLUCK_PATH']
 
-        extension = cls(filename=filename)
+        extension = cls(directory=directory)
         crawler.signals.connect(extension.item_scraped, signal=signals.item_scraped)
         crawler.signals.connect(extension.spider_closed, signal=signals.spider_closed)
 
@@ -33,15 +33,18 @@ def item_scraped(self, item, spider):
             return
 
         self.spiders_seen.add(spider.name)
-        with open(self.filename, 'a+') as output:
-            output.write(f"{item['value']},{spider.name}\n")
+
+        self._write(spider, item['value'])
 
     def spider_closed(self, spider, reason):
         if not spider.pluck or spider.name in self.spiders_seen:
             return
 
-        with open(self.filename, 'a+') as output:
-            output.write(f"{reason},{spider.name}\n")
+        self._write(spider, reason)
+
+    def _write(self, spider, value):
+        with open(os.path.join(self.directory, _pluck_filename(spider)), 'a+') as f:
+            f.write(f'{value},{spider.name}\n')
 
 
 class KingfisherFilesStore:

kingfisher_scrapy/util.py

@@ -8,6 +8,15 @@
 from ijson import ObjectBuilder, utils
 
 
+def _pluck_filename(opts):
+    if opts.package_pointer:
+        parts = ['pluck', 'package', opts.package_pointer[1:].replace('/', '-')]
+    else:
+        parts = ['pluck', 'release', opts.release_pointer[1:].replace('/', '-')]
+
+    return f"{'-'.join(parts)}.csv"
+
+
 def components(start, stop=None):
     """
     Returns a function that returns the selected non-empty path components, excluding the ``.json`` extension.

tests/extensions/test_kingfisher_pluck.py

@@ -1,4 +1,5 @@
 import os
+from glob import glob
 from tempfile import TemporaryDirectory
 
 from kingfisher_scrapy.extensions import KingfisherPluck
@@ -17,7 +18,7 @@ def test_disabled():
         extension.item_scraped(item, spider)
         extension.spider_closed(spider, 'itemcount')
 
-        assert not os.path.exists(os.path.join(tmpdirname, 'pluck.csv'))
+        assert not glob(os.path.join(tmpdirname, 'pluck*.csv'))
 
 
 def test_item_scraped():
@@ -30,21 +31,21 @@ def test_item_scraped():
 
         extension.item_scraped(item, spider)
 
-        with open(os.path.join(tmpdirname, 'pluck.csv')) as f:
+        with open(os.path.join(tmpdirname, 'pluck-release-date.csv')) as f:
             assert '2020-10-01,test\n' == f.read()
 
         # Only one item from the same spider is written.
         extension.item_scraped(item, spider)
 
-        with open(os.path.join(tmpdirname, 'pluck.csv')) as f:
+        with open(os.path.join(tmpdirname, 'pluck-release-date.csv')) as f:
             assert '2020-10-01,test\n' == f.read()
 
         # An item from another spider is appended.
         spider.name = 'other'
         item['value'] = '2020-10-02'
         extension.item_scraped(item, spider)
 
-        with open(os.path.join(tmpdirname, 'pluck.csv')) as f:
+        with open(os.path.join(tmpdirname, 'pluck-release-date.csv')) as f:
             assert '2020-10-01,test\n2020-10-02,other\n' == f.read()
 
 
@@ -59,7 +60,7 @@ def test_spider_closed_with_items():
         extension.item_scraped(item, spider)
         extension.spider_closed(spider, 'itemcount')
 
-        with open(os.path.join(tmpdirname, 'pluck.csv')) as f:
+        with open(os.path.join(tmpdirname, 'pluck-release-date.csv')) as f:
             assert '2020-10-01,test\n' == f.read()
 
 
@@ -72,5 +73,5 @@ def test_spider_closed_without_items():
 
         extension.spider_closed(spider, 'itemcount')
 
-        with open(os.path.join(tmpdirname, 'pluck.csv')) as f:
+        with open(os.path.join(tmpdirname, 'pluck-release-date.csv')) as f:
             assert 'itemcount,test\n' == f.read()