create reverse dfa b4 reassigning name

README.md

@@ -1,4 +1,5 @@
 [WIP]
+
 # full_zk_regex
 
 Slide: https://docs.google.com/presentation/d/1nSZdmwDKXjEM6bP6WBYyAWbCgK4cjpm-SXqDAA-MOjE/edit?usp=sharing
@@ -24,27 +25,28 @@ Data flow and related functions
 2. Frontend process this regex, by simp_regex = gen_dfa.simplifyRegex
    (to deal with escape and change [] into or statements), then let users high light start and end of each group (inclusive) of this simp_regex. Save the value as submatches e.g.
    const submatches = [
-   [5, 30],
+   [5, 29],
    [7, 13],
-   [15, 25],
+   [15, 24],
    ];
 3. Match all the strings that satisfies regex as a whole. ฺ(like previously version)
    Run
    const simp_graph = gen_dfa.simplifyGraph(regex);
-   
+
    const matched_dfa = gen_dfa.findSubstrings(simp_graph, text);
-   
+
    for (const subs of matched_dfa[1]) {
-   
+
    var matched = text.slice(subs[0], subs[1] + 1);} Very important of +1!!
+
 4. Now for each matched, we start extract substring state (we can have multiple matched values!)
 
    const tagged_simp_graph = gen_tagged_dfa.tagged_simplifyGraph(regex, submatches);
-   
+
    var final_graph = gen_tagged_dfa.findMatchStateTagged(tagged_simp_graph);
-   
+
    var allTags = final_graph["tags"];
-   
+
    var transitions = final_graph["transitions"];
 
 5. For Circom, we run

src/gen.js

@@ -251,7 +251,7 @@ export function regexToM1(text, submatches) {
   }
 
   // New: simplifyRegex and simplify Plus
-  console.log("herer in Gen: ", simplifyRegex(text));
+  // console.log("simplify regex in Gen: ", simplifyRegex(text));
   let after_plus = simplifyPlus(simplifyRegex(text), submatches);
 
   // console.log("afterrr; ", after_plus["submatches"]);

src/gen_rev_dfa.js

@@ -0,0 +1,76 @@
+// given simp_graph (plain dfa), return its reversed version.
+export function reverseDFA(simp_graph) {
+  // simp_q = [
+  //     '0',  '1', '2',  '3',
+  //     '4',  '5', '6',  '7',
+  //     '8',  '9', '10', '11',
+  //     '12'
+  //   ]
+  var simp_q = simp_graph["states"];
+  var simp_transition = simp_graph["transitions"];
+  var rev_q = [];
+  // set rev_q to [{f}]
+  rev_q.push(simp_graph["accepted_states"]);
+  var rev_q_all = new Set();
+  var rev_transition = {};
+  var visited = new Set();
+  var simp_start = new Set();
+  simp_start.add(simp_graph["start_state"]);
+  var rev_accepted = new Set();
+  var rev_start = simp_graph["accepted_states"];
+
+  // inside loop
+  while (rev_q.length > 0) {
+    var state_set = rev_q.pop();
+    var states_id = [];
+    for (const state of state_set) {
+      states_id.push(parseInt(state));
+    }
+    states_id.sort((a, b) => a - b);
+    states_id = states_id.toString();
+    if (visited.has(states_id)) {
+      continue;
+    }
+    var checkStart = states_id.split(",");
+    for (const state of checkStart) {
+      if (simp_start.has(state)) {
+        rev_accepted.add(states_id);
+        break;
+      }
+    }
+    rev_q_all.add(states_id);
+    visited.add(states_id);
+    var alp_dict = {};
+    for (let from in simp_transition) {
+      for (let alphabet in simp_transition[from]) {
+        if (state_set.has(simp_transition[from][alphabet])) {
+          if (!alp_dict.hasOwnProperty(alphabet)) {
+            alp_dict[alphabet] = new Set();
+          }
+          alp_dict[alphabet].add(from);
+        }
+      }
+    }
+    for (let alp in alp_dict) {
+      if (alp_dict[alp].size > 0) {
+        rev_q.push(alp_dict[alp]);
+        var alp_string = [];
+        for (const state of alp_dict[alp]) {
+          alp_string.push(parseInt(state));
+        }
+        alp_string.sort((a, b) => a - b);
+        alp_string = alp_string.toString();
+        if (!rev_transition.hasOwnProperty(states_id)) {
+          rev_transition[states_id] = {};
+        }
+        rev_transition[states_id][alp] = alp_string;
+      }
+    }
+  }
+  return {
+    states: Array.from(rev_q_all),
+    start_state: rev_start,
+    accept_states: rev_accepted,
+    transitions: rev_transition,
+  };
+}

src/test.js

@@ -1,14 +1,37 @@
-import { simplifyGraph, findSubstrings } from "./gen_dfa";
+import { simplifyGraph, findSubstrings, simplifyRegex } from "./gen_dfa";
+import {
+  tagged_simplifyGraph,
+  findMatchStateTagged,
+  formatForCircom,
+} from "./gen_tagged_dfa";
+import { reverseDFA } from "./gen_rev_dfa";
 function test() {
   const text =
     "adsfasd DKI: v=12/; d=22; a=//121; d=1; bh=xUqTs2T2FPGCOB52 sdflj";
   const regex = "DKI: (([vad]=([12/]+); )+)bh";
+  const simp_regex = simplifyRegex(regex);
+  console.log("simp_regex: ", simp_regex);
+  const submatches = [
+    [5, 29],
+    [7, 13],
+    [15, 24],
+  ];
   const simp_graph = simplifyGraph(regex);
+  console.log("simp graph: ", simp_graph);
+  const rev_graph = reverseDFA(simp_graph);
+  console.log("rev graph: ", rev_graph);
   const matched_dfa = findSubstrings(simp_graph, text);
-  for (const subs of matched_dfa[1]) {
-    var matched = text.slice(subs[0], subs[1] + 1);
-    console.log("matched: ", matched);
-  }
+  // for (const subs of matched_dfa[1]) {
+  //   var matched = text.slice(subs[0], subs[1] + 1);
+  //   console.log("matched: ", matched);
+  // }
+  const tagged_simp_graph = tagged_simplifyGraph(regex, submatches);
+  var final_graph = findMatchStateTagged(tagged_simp_graph);
+  var allTags = final_graph["tags"];
+  var transitions = final_graph["transitions"];
+  console.log("final graph: ", final_graph);
+  var circom_graph = formatForCircom(final_graph);
+  var circom_rev_graph = formatForCircom(rev_graph);
 }
 describe("test backend", function () {
   it("should print correctly", function () {