From 17778ac3ab8598eb3d43f562a092b9aa7c0a1a42 Mon Sep 17 00:00:00 2001
From: Star Rauchenberger <fefferburbia@gmail.com>
Date: Sat, 2 Dec 2023 17:10:48 -0500
Subject: Filter out profane words

---
 generator/generator.cpp | 27 +++++++++++++++++++++++----
 1 file changed, 23 insertions(+), 4 deletions(-)

(limited to 'generator/generator.cpp')

diff --git a/generator/generator.cpp b/generator/generator.cpp
index 7ab69b5..0309482 100644
--- a/generator/generator.cpp
+++ b/generator/generator.cpp
@@ -4,6 +4,7 @@
 #include <hkutil/string.h>
 
 #include <algorithm>
+#include <filesystem>
 #include <fstream>
 #include <list>
 #include <regex>
@@ -11,6 +12,7 @@
 #include <stdexcept>
 #include <string>
 #include <unordered_map>
+#include <unordered_set>
 #include <vector>
 
 constexpr int MIN_FREQUENCY = 2000000;
@@ -49,11 +51,12 @@ std::list<std::string> readFile(std::string path, bool uniq = false) {
 
 generator::generator(std::string agidPath, std::string wordNetPath,
                      std::string cmudictPath, std::string wordfreqPath,
-                     std::string outputPath)
+                     std::string datadirPath, std::string outputPath)
     : agidPath_(agidPath),
       wordNetPath_(wordNetPath),
       cmudictPath_(cmudictPath),
       wordfreqPath_(wordfreqPath),
+      datadirPath_(datadirPath),
       outputPath_(outputPath) {
   // Ensure AGID infl.txt exists
   if (!std::ifstream(agidPath_)) {
@@ -102,6 +105,14 @@ void generator::run() {
     }
   }
 
+  std::unordered_set<std::string> profane;
+  {
+    std::list<std::string> lines(readFile(datadirPath_ / "profane.txt"));
+    for (const std::string& line : lines) {
+      profane.insert(line);
+    }
+  }
+
   {
     std::list<std::string> lines(readFile(wordNetPath_ + "wn_s.pl"));
     hatkirby::progress ppgs("Reading synsets from WordNet...", lines.size());
@@ -142,6 +153,11 @@ void generator::run() {
         continue;
       }
 
+      // Ignore any profane words.
+      if (profane.count(text)) {
+        continue;
+      }
+
       // The WordNet data does contain duplicates, so we need to check that we
       // haven't already created this word.
       std::pair<int, int> lookup(synset_id, wnum);
@@ -175,7 +191,8 @@ void generator::run() {
       }
 
       if (!word_by_base_.count(infinitive) &&
-          !(type == 'V' && word_frequencies[infinitive] >= MIN_FREQUENCY)) {
+          !(type == 'V' && word_frequencies[infinitive] >= MIN_FREQUENCY &&
+            !profane.count(infinitive))) {
         continue;
       }
 
@@ -262,8 +279,10 @@ void generator::run() {
       // Compile the forms we have mapped.
       for (const std::list<std::string>& infl_list : inflections) {
         for (const std::string& infl : infl_list) {
-          size_t form_id = LookupOrCreateForm(infl);
-          AddFormToWord(form_id, word_id);
+          if (!profane.count(infl)) {
+            size_t form_id = LookupOrCreateForm(infl);
+            AddFormToWord(form_id, word_id);
+          }
         }
       }
     }
-- 
cgit 1.4.1