EntryDiff scoring (#4)

* Adds EntryDiff.score calculation * Update .travis.yml * Updates issues * Fixes http tx dict product * Revert "Update .travis.yml" This reverts commit a64a6e1. * Adds UI * Adds labels coloring * Adds decimalAdjust * Moves score to Comparison * Adds soft diffs into scoring result * Removes TODO Co-authored-by: Andrey Pokhilko <[email protected]>
up9inc · Sep 14, 2021 · 04ce196 · 04ce196
1 parent 4ab40b2
commit 04ce196
Show file tree

Hide file tree

Showing 13 changed files with 226 additions and 36 deletions.
diff --git a/Dockerfile b/Dockerfile
@@ -24,6 +24,7 @@ ENV SPA_LOCATION=/app/harnic-spa
 # install dependencies
 COPY requirements.txt .
 RUN pip install -r requirements.txt
+RUN python -m spacy download en_core_web_sm
 
 # install backend
 COPY harnic harnic

diff --git a/README.md b/README.md
@@ -64,3 +64,5 @@ docker rm harnic
 - Add contextual context-wrap (*Hard*)
 - Add request postData
 - Handle soft diffs missing or added as soft (*Under question*)
+- Http tx scores to 0.5/0.5
+- Add score of entry and body to UI
diff --git a/harnic-spa/src/App.css b/harnic-spa/src/App.css
@@ -24,16 +24,31 @@
   background-color: rgba(0, 0, 0, 0.25)
 }
 */
-.ui.table td.warning, .ui.table tr.warning {
-    background: rgba(246, 157, 80, 0.2) !important;
+.ui.table td.warning, .ui.table tr.warning, .label.warning {
+  background: rgba(246, 157, 80, 0.2) !important;
 }
 
-.ui.table td.positive, .ui.table tr.positive {
-    background: rgba(70, 149, 74, 0.2) !important;
+.ui.table td.positive, .ui.table tr.positive, .label.positive {
+  background: rgba(70, 149, 74, 0.2) !important;
 }
 
-.ui.table td.negative, .ui.table tr.negative {
-    background: rgba(201, 60, 55, 0.2) !important;
+.ui.table td.negative, .ui.table tr.negative, .label.negative {
+  background: rgba(201, 60, 55, 0.2) !important;
+}
+
+.label.warning {
+  background: rgba(246, 157, 80, 0.55) !important;
+  color: white !important;
+}
+
+.label.positive {
+  background: rgba(70, 149, 74, 0.55) !important;
+  color: white !important;
+}
+
+.label.negative {
+  background: rgba(201, 60, 55, 0.55) !important;
+  color: white !important;
 }
 
 /* Some basic formatting */
@@ -180,14 +195,21 @@ code {
 }
 
 .reordering-icon {
-  float: right;
   -webkit-transform: rotate(90deg);
   -moz-transform: rotate(90deg);
   -ms-transform: rotate(90deg);
   -o-transform: rotate(90deg);
   transform: rotate(90deg);
 }
 
+.entry-diff.meta-label {
+  float: right;  
+}
+
+.entry-diff.meta-label.reordering {
+  padding-right: 1px !important;  
+}
+
 .truncated {
   text-align: center;
   font-family: 'Lato', auto;

diff --git a/harnic-spa/src/components/DiffRecordRow.js b/harnic-spa/src/components/DiffRecordRow.js
@@ -1,9 +1,9 @@
 import { useState } from "react";
-import { Tab, Table, Icon, Label } from "semantic-ui-react";
+import { Tab, Table, Icon, Label, Menu } from "semantic-ui-react";
 
 import RequestData from "./RequestData.js";
 import ResponseData from "./ResponseData.js";
-import { truncate } from ".././utils.js";
+import { truncate, getScoreLabelClass, decimalAdjust } from ".././utils.js";
 
 const DiffRecordRow = ({ record }) => {
   const [isOpen, setIsOpen] = useState(false);
@@ -31,7 +31,16 @@ const DiffRecordRow = ({ record }) => {
   const aPanes = record.pair.a
     ? [
         {
-          menuItem: "Request",
+          menuItem: (
+            <Menu.Item key='request'>
+              Request
+              {record.tag === "diff" &&
+                <Label className={getScoreLabelClass(record.diff.score.by_http_tx_type.request)} floating>
+                  {decimalAdjust('floor', record.diff.score.by_http_tx_type.request * 100, -1)}%
+                </Label>
+              }
+            </Menu.Item>
+          ),
           render: () => (
             <Tab.Pane>
               <RequestData
@@ -42,7 +51,16 @@ const DiffRecordRow = ({ record }) => {
           ),
         },
         {
-          menuItem: "Response",
+          menuItem: (
+            <Menu.Item key='response'>
+              Response
+              {record.tag === "diff" &&
+                <Label className={getScoreLabelClass(record.diff.score.by_http_tx_type.response)} floating>
+                  {decimalAdjust('floor', record.diff.score.by_http_tx_type.response * 100, -1)}%
+                </Label>
+              }
+            </Menu.Item>
+          ),
           render: () => (
             <Tab.Pane>
               <ResponseData
@@ -51,6 +69,7 @@ const DiffRecordRow = ({ record }) => {
                 response={record.pair.a.response}
                 diff={record.diff && record.diff.comparisons.response}
                 initialEntry={true}
+                score={record.diff && record.diff.score}
               />
             </Tab.Pane>
           ),
@@ -69,7 +88,16 @@ const DiffRecordRow = ({ record }) => {
   const bPanes = record.pair.b
     ? [
         {
-          menuItem: "Request",
+          menuItem: (
+            <Menu.Item key='request'>
+              Request
+              {record.tag === "diff" &&
+                <Label className={getScoreLabelClass(record.diff.score.by_http_tx_type.request)} floating>
+                  {decimalAdjust('floor', record.diff.score.by_http_tx_type.request * 100, -1)}%
+                </Label>
+              }
+            </Menu.Item>
+          ),
           render: () => (
             <Tab.Pane>
               <RequestData
@@ -80,7 +108,16 @@ const DiffRecordRow = ({ record }) => {
           ),
         },
         {
-          menuItem: "Response",
+          menuItem: (
+            <Menu.Item key='response'>
+              Response
+              {record.tag === "diff" &&
+                <Label className={getScoreLabelClass(record.diff.score.by_http_tx_type.response)} floating>
+                  {decimalAdjust('floor', record.diff.score.by_http_tx_type.response * 100, -1)}%
+                </Label>
+              }
+            </Menu.Item>
+          ),
           render: () => (
             <Tab.Pane>
               <ResponseData
@@ -89,6 +126,7 @@ const DiffRecordRow = ({ record }) => {
                 response={record.pair.b.response}
                 diff={record.diff && record.diff.comparisons.response}
                 initialEntry={false}
+                score={record.diff && record.diff.score}
               />
             </Tab.Pane>
           ),
@@ -124,8 +162,15 @@ const DiffRecordRow = ({ record }) => {
             </>
           )}
           {record.is_reordering && (
-            <Icon name="exchange" className="reordering-icon" />
+            <Label className="entry-diff meta-label reordering">
+              <Icon name="exchange" className="reordering-icon" />
+            </Label>
           )}
+          {record.tag === "diff" &&
+            <Label className="entry-diff meta-label">
+              {decimalAdjust('floor', record.diff.score.final * 100, -1)}%
+            </Label>
+          }
         </Table.Cell>
         <Table.Cell>
           {record.pair.b && (

diff --git a/harnic-spa/src/components/ResponseData.js b/harnic-spa/src/components/ResponseData.js
@@ -11,7 +11,7 @@ import { DateTime } from "luxon";
 import regexifyString from "regexify-string";
 
 import ModalScrollingContent from "./ModalScrollingContent.js";
-import { truncate, calculateDiffClass } from ".././utils.js";
+import { truncate, calculateDiffClass, getScoreLabelClass, decimalAdjust } from ".././utils.js";
 
 const ContentText = ({ value, request }) => {
   if (value === null) {
@@ -69,6 +69,7 @@ const ResponseData = ({
   response,
   diff,
   initialEntry,
+  score,
 }) => {
   const cmpIdx = initialEntry ? 0 : 1;
 
@@ -274,6 +275,11 @@ const ResponseData = ({
         <List.Item>
           <div>
             <b>Content:</b>
+            {score && 
+                <Label className={getScoreLabelClass(score.full.response.content)}>
+                  {decimalAdjust('floor', score.full.response.content * 100, -1)}%
+                </Label>
+            }            
           </div>
           <List>
             {Object.entries(response.content).map(([key, value]) => {

diff --git a/harnic-spa/src/utils.js b/harnic-spa/src/utils.js
@@ -20,5 +20,34 @@ const calculateDiffClass = (diff, criteria, key) => {
   return keyClass;
 };
 
- export { truncate, calculateDiffClass };
+const getScoreLabelClass = (score) => {
+  if (score < 0.25) {
+    return 'negative';
+  } else if (score == 1) {
+    return 'positive';
+  } else {
+    return 'warning';
+  }
+};
+
+const decimalAdjust = (type, value, exp) => {
+  // If the exp is undefined or zero...
+  if (typeof exp === 'undefined' || +exp === 0) {
+    return Math[type](value);
+  }
+  value = +value;
+  exp = +exp;
+  // If the value is not a number or the exp is not an integer...
+  if (isNaN(value) || !(typeof exp === 'number' && exp % 1 === 0)) {
+    return NaN;
+  }
+  // Shift
+  value = value.toString().split('e');
+  value = Math[type](+(value[0] + 'e' + (value[1] ? (+value[1] - exp) : -exp)));
+  // Shift back
+  value = value.toString().split('e');
+  return +(value[0] + 'e' + (value[1] ? (+value[1] + exp) : exp));
+}
+
+export { truncate, calculateDiffClass, getScoreLabelClass, decimalAdjust };
 
diff --git a/harnic/compare/entry.py b/harnic/compare/entry.py
@@ -1,7 +1,7 @@
 from functools import partial
 
-from harnic.compare.utils import dict_compare, qp_compare, scalars_compare, content_compare
-from harnic.constants import SOFT_HEADER_KEYS
+from harnic.compare.utils import content_compare, dict_compare, dict_product, qp_compare, scalars_compare
+from harnic.constants import SCORE_COEFS, SCORE_HTTP_TX_TYPE_COEFS, SOFT_HEADER_KEYS
 
 headers_compare = partial(dict_compare, exceptions=SOFT_HEADER_KEYS, exculde_values=True)
 
@@ -13,25 +13,52 @@ def __init__(self, a, b):
         self.a = a
         self.b = b
         self.equal = None
+        self.score = {}
         self.comparisons = self._get_diff()
 
     def _get_diff(self):
         # method and url are not handled here as long as they are part of the entry hash
         comparisons = {'request': {}, 'response': {}}
 
-        comparisons['request']['bodySize'] = scalars_compare(self.a.request['bodySize'],
-                                                             self.b.request['bodySize'])
-        comparisons['request']['query_params'] = qp_compare(self.a.request['url'].query_params,
-                                                              self.b.request['url'].query_params)
-        comparisons['request']['headers'] = headers_compare(self.a.request['headers'],
-                                                            self.b.request['headers'])
+        diff_score = {
+            'request': {},
+            'response': {},
+        }
 
-        comparisons['response']['status'] = scalars_compare(self.a.response['status'],
-                                                            self.b.response['status'])
-        comparisons['response']['headers'] = headers_compare(self.a.response['headers'],
-                                                             self.b.response['headers'])
-        comparisons['response']['content'] = content_compare(self.a.response,
-                                                             self.b.response)
+        cmp = scalars_compare(self.a.request['url'].url, self.b.request['url'].url)
+        diff_score['request']['url'] = cmp.score
+
+        cmp = scalars_compare(self.a.request['bodySize'], self.b.request['bodySize'])
+        comparisons['request']['bodySize'] = cmp
+
+        cmp = qp_compare(self.a.request['url'].query_params, self.b.request['url'].query_params)
+        comparisons['request']['query_params'], diff_score['request']['query_params'] = cmp, cmp.score
+
+        cmp = headers_compare(self.a.request['headers'], self.b.request['headers'])
+        comparisons['request']['headers'], diff_score['request']['headers'] = cmp, cmp.score
+
+        # TODO: implement postData cmp
+        diff_score['request']['postData'] = 1  # Treat same for now
+
+        cmp = scalars_compare(self.a.response['status'], self.b.response['status'])
+        comparisons['response']['status'], diff_score['response']['status'] = cmp, cmp.score
+
+        cmp = headers_compare(self.a.response['headers'], self.b.response['headers'])
+        comparisons['response']['headers'], diff_score['response']['headers'] = cmp, cmp.score
+
+        cmp = content_compare(self.a.response, self.b.response)
+        comparisons['response']['content'], diff_score['response']['content'] = cmp, cmp.score
 
         self.equal = all(all(cmp.equal for cmp in criteria.values()) for criteria in comparisons.values())
+
+        self.score['full'] = diff_score
+        diff_score_with_coefs = {
+            'request': sum(dict_product(diff_score['request'], SCORE_COEFS['request']).values()),
+            'response': sum(dict_product(diff_score['response'], SCORE_COEFS['response']).values()),
+        }
+        self.score['by_http_tx_type'] = diff_score_with_coefs
+        diff_score_with_coefs = dict_product(diff_score_with_coefs, SCORE_HTTP_TX_TYPE_COEFS)
+        final_score = sum(diff_score_with_coefs.values())
+        self.score['final'] = final_score
+
         return comparisons
diff --git a/harnic/compare/matcher.py b/harnic/compare/matcher.py
@@ -64,6 +64,7 @@ def _build_files_diff(opcodes, file1, file2):
         PermTag.DIFF: 0,
         PermTag.INSERT: 0,
         PermTag.DELETE: 0,
+        '_diff_scores_sum': 0,
     }
     perms_total = _calculate_permutations_total_number(opcodes)
     with tqdm(total=perms_total, desc='Constructing diff records') as pbar:
@@ -77,6 +78,8 @@ def _build_files_diff(opcodes, file1, file2):
                     dr = DiffRecord(pair, entry_diff, tag_selector)
                     records.append(dr)
                     stats[tag_selector] += 1
+                    if tag_selector == PermTag.DIFF:
+                        stats['_diff_scores_sum'] += entry_diff.score['final']
                     pbar.update()
             elif tag == 'replace':
                 for i in range(i1, i2):
@@ -101,15 +104,15 @@ def _build_files_diff(opcodes, file1, file2):
                     records.append(dr)
                     stats[PermTag.INSERT] += 1
                     pbar.update()
-            stats['ratio'] = 2.0 * stats[PermTag.EQUAL] / (len(file1) + len(file2))
+
+    stats['ratio'] = 2.0 * (stats[PermTag.EQUAL] + stats['_diff_scores_sum']) / (len(file1) + len(file2))
 
     reorders = _calculate_reorders(records)
     reorders_stats = _calculate_reorders_stats(reorders, stats, (file1, file2))
     stats = {
         'with_reorders': reorders_stats,
         'strict_order': stats,
     }
-
     return records, reorders, stats
 
 
@@ -162,7 +165,9 @@ def _calculate_reorders_stats(reorders, stats, files):
     for reorder in reorders:
         tag_selector = PermTag.EQUAL if reorder['entry_diff'].equal else PermTag.DIFF
         stats[tag_selector] += 1
-    stats['ratio'] = 2.0 * stats[PermTag.EQUAL] / (len(file1) + len(file2))
+        if tag_selector == PermTag.DIFF:
+            stats['_diff_scores_sum'] += reorder['entry_diff'].score['final']
+    stats['ratio'] = 2.0 * (stats[PermTag.EQUAL] + stats['_diff_scores_sum']) / (len(file1) + len(file2))
 
     return stats
 

diff --git a/harnic/compare/schemas.py b/harnic/compare/schemas.py
@@ -19,6 +19,7 @@ class ComparisonSchema(Schema):
 class EntryDiffSchema(Schema):
     equal = fields.Bool()
     comparisons = fields.Dict(values=fields.Dict(values=fields.Nested('ComparisonSchema')))
+    score = fields.Dict()
 
 
 class PairSchema(Schema):