ease/document extra dependency

capa/analysis/requirements.txt

@@ -0,0 +1,2 @@
+# temporary extra file to track dependencies of the analysis directory
+nltk==3.9.1

capa/analysis/strings/__init__.py

@@ -133,6 +133,7 @@ def prune_databases(dbs: list[LibraryStringDatabase], n=8):
 
     try:
         from nltk.corpus import words as nltk_words
+        nltk_words.words()
     except ImportError:
         # one-time download of dataset.
         # this probably doesn't work well for embedded use.