summaryrefslogtreecommitdiff
path: root/priv/irc/txt/markov.py
diff options
context:
space:
mode:
authorhref <href@random.sh>2020-03-11 21:18:34 +0100
committerhref <href@random.sh>2020-03-11 21:18:34 +0100
commita28d24470ddeca6196219a1333c1ccac1319efef (patch)
tree4f29e3c8fb6afbb1f99d6b8737f844c95fca54df /priv/irc/txt/markov.py
parentup to 420*100 (diff)
welp
Diffstat (limited to '')
-rw-r--r--priv/irc/txt/markov.py27
1 files changed, 27 insertions, 0 deletions
diff --git a/priv/irc/txt/markov.py b/priv/irc/txt/markov.py
new file mode 100644
index 0000000..b72016e
--- /dev/null
+++ b/priv/irc/txt/markov.py
@@ -0,0 +1,27 @@
+import markovify
+import os
+import string
+import sys
+import shlex
+import unidecode
+
+combined_model = None
+text = ""
+
+dir = sys.argv[1]
+
+for (dirpath, _, filenames) in os.walk(dir):
+ for filename in filenames:
+ with open(os.path.join(dirpath, filename)) as f:
+ text = text + unidecode.unidecode(f.read().lower())
+
+
+# Build the model.
+text_model = markovify.NewlineText(text, well_formed = False)
+
+if len(sys.argv) > 1:
+ insp = " ".join(map(shlex.quote, sys.argv[2:])).lower()
+ print(text_model.make_sentence_with_start(unidecode.unidecode(insp), strict = False, tries = 1000))
+else:
+ print(text_model.make_short_sentence(280, tries=1000))
+