Sankey syntax has beed reduced

cypress/integration/rendering/sankey.spec.js

@@ -5,6 +5,7 @@ describe('Sankey Diagram', () => {
     imgSnapshotTest(
       `
       sankey
+      
       a,b,10
       `,
       {}

packages/mermaid/src/diagrams/sankey/parser/energy.csv

@@ -1,5 +1,7 @@
-Agricultural 'waste',Bio-conversion,124.729
+%% there are leading and trailing spaces, do not remove
+    Agricultural 'waste',Bio-conversion,124.729   
 Bio-conversion,Liquid,0.597
+%% line with comment
 Bio-conversion,Losses,26.862
 Bio-conversion,Solid,280.322
 Bio-conversion,Gas,81.144
@@ -50,6 +52,9 @@ Oil reserves,Oil,107.703
 Oil,Liquid,611.99
 Other waste,Solid,56.587
 Other waste,Bio-conversion,77.81
+%% blank lines in the middle
+
+
 Pumped heat,Heating and cooling - homes,193.026
 Pumped heat,Heating and cooling - commercial,70.672
 Solar PV,Electricity grid,59.901
@@ -65,4 +70,7 @@ Thermal generation,District heating,79.329
 Tidal,Electricity grid,9.452
 UK land based bioenergy,Bio-conversion,182.01
 Wave,Electricity grid,19.013
-Wind,Electricity grid,289.366
+Wind,Electricity grid,289.366
+
+%% lines at the end, do not remove
+

packages/mermaid/src/diagrams/sankey/parser/sankey.jison

@@ -23,10 +23,10 @@ TEXTDATA [\u0020-\u0021\u0023-\u002B\u002D-\u007E]
 
 %%
 
-<<EOF>> { return 'EOF' }
+<<EOF>> { return 'EOF' } // match end of file
 
 "sankey" { return 'SANKEY' }
-({CRLF}|{LF})+ { return 'NEWLINE' } // let newline to be multiple lines
+({CRLF}|{LF}) { return 'NEWLINE' }
 {COMMA} { return 'COMMA' }
 {DQUOTE} { return 'DQUOTE' }
 {TEXTDATA}* { return 'NON_ESCAPED_TEXT' }
@@ -38,28 +38,11 @@ TEXTDATA [\u0020-\u0021\u0023-\u002B\u002D-\u007E]
 
 %% // language grammar
 
-start
+start: SANKEY NEWLINE csv opt_eof;
-  : SANKEY csv opt_eof
-  ;
 
-csv
+csv: record csv_tail;
-  : record csv_tail 
+csv_tail: NEWLINE csv | ;
-  ;
+opt_eof: EOF | ;
-
-csv_tail
-  : NEWLINE csv
-  | // empty
-  ;
-
-opt_newline
-  : NEWLINE
-  | // empty
-  ;
-
-opt_eof
-  : EOF
-  | // empty
-  ;
 
 record
   : field\[source] COMMA field\[target] COMMA field\[value] {
@@ -68,7 +51,6 @@ record
       const value = parseFloat($value.trim());
       $$ = yy.addLink(source,target,value);
     } // parse only 3 fields, this is not part of CSV standard
-  | {} // allow empty record to handle empty lines, this is not part of CSV standard either
   ;
 
 field

packages/mermaid/src/diagrams/sankey/parser/sankey.spec.ts

@@ -3,7 +3,8 @@ import diagram from './sankey.jison';
 // @ts-ignore: jison doesn't export types
 import { parser } from './sankey.jison';
 import db from '../sankeyDB.js';
-// import { fail } from 'assert';
+import { cleanupComments } from '../../../diagram-api/comments.js';
+import { prepareTextForParsing } from '../sankeyDiagram.js';
 
 describe('Sankey diagram', function () {
   // TODO - these examples should be put into ./parser/stateDiagram.spec.js
@@ -18,15 +19,16 @@ describe('Sankey diagram', function () {
       const fs = await import('fs');
       const path = await import('path');
       const csv = path.resolve(__dirname, './energy.csv');
-      fs.readFile(csv, 'utf8', (err: NodeJS.ErrnoException | null, data: string) => {
+      const data = fs.readFileSync(csv, 'utf8');
-        if (err) {
-          throw err;
-        }
 
-        const str = `sankey\\n${data}`;
+      // Add \n\n + space to emulate possible possible imperfections
+      const graphDefinition = prepareTextForParsing(cleanupComments('sankey\n\n ' + data));
+      // const textToParse = graphDefinition
+      //   .replaceAll(/^[^\S\r\n]+|[^\S\r\n]+$/g, '')
+      //   .replaceAll(/([\n\r])+/g, "\n")
+      //   .trim();
 
-        parser.parse(str);
+      parser.parse(graphDefinition);
-      });
     });
   });
 });

packages/mermaid/src/diagrams/sankey/sankeyDiagram.ts

@@ -5,6 +5,18 @@ import db from './sankeyDB.js';
 import styles from './styles.js';
 import renderer from './sankeyRenderer.js';
 
+export const prepareTextForParsing = (text: string): string => {
+  const textToParse = text
+    .replaceAll(/^[^\S\r\n]+|[^\S\r\n]+$/g, '') // remove all trailing spaces for each row
+    .replaceAll(/([\n\r])+/g, '\n') // remove empty lines duplicated
+    .trim();
+
+  return textToParse;
+};
+
+const originalParse = parser.parse.bind(parser);
+parser.parse = (text: string) => originalParse(prepareTextForParsing(text));
+
 export const diagram: DiagramDefinition = {
   parser,
   db,