Angular Material Table rowspan columns based on da

Question

Even now in version 7.2 of Angular Material, I can't seem to find examples on how to use rowspan on mat-table and keep the component functionality.

This is how far (short?) I've got:

https://stackblitz.com/edit/angular-wudscb

The example in the Stackblitz above is "almost" what I am looking for, but I am not being able to see how to finish it.

...
===============================================
||     ||            ||            ||  row1  ||
||  1  ||  Hydrogen  ||   1.0079   ||========||
||     ||            ||            ||  row2  ||
===============================================
||     ||            ||            ||  row1  ||
||     ||            ||            ||========||
||  2  ||   Helium   ||   4.0026   ||  row2  ||
||     ||            ||            ||========||
||     ||            ||            ||  row3  ||
===============================================
||     ||            ||            ||  row1  ||
||  3  ||  Lithium   ||   6.941    ||========||
||     ||            ||            ||  row2  ||
===============================================
...

An example using other metadata format can be found in:

https://stackblitz.com/edit/angular-lnahlh

Following my Stackblitz (the first link), my questions are:

Am I too far of achieving this rowspan shim/hack?

How do I loop the rows based on the lenght of the row['descriptions'] size?

What If I had another array property inside the object? Could I iterate and generate the columns/rows/rowspan with its size, so it would become more generic?

I'm trying to find a generic solution for the community.

Answer 1

As I was not happy with the answers presented (especially with the cacheSpan implementation).

I came up with something more convenient IMHO, that goes something like:

export class TableBasicExample {
  displayedColumns = ['priority', 'status', 'dateCreated', 'testNumber', 'testCurrency', 'testTime'];
  dataSource = DATA;

  spans = {};

  constructor() {
    this.spans = Object.assign({}, {
      priority: this.spanDeep(['priority'], DATA),
      status: this.spanDeep(['priority', 'status'], DATA),
      dateCreated: this.spanDeep(['priority', 'status', 'dateCreated'], DATA)
    });
  }

  spanDeep(paths: string[] | null, data: any[]) {
    if (!paths.length) {
      return [...data]
        .fill(0)
        .fill(data.length, 0, 1);
    }

    const copyPaths = [...paths];
    const path = copyPaths.shift();

    const uniq = uniqWith(data, (a, b) => get(a, path) === get(b, path))
      .map(item => get(item, path));

    return uniq
      .map(uniqItem => this.spanDeep(copyPaths, data.filter(item => uniqItem === get(item, path))))
      .flat(paths.length);
  }

  getRowSpan(path, idx) {
    return this.spans[path][idx];
  }
};

Working example may be found here: https://stackblitz.com/edit/angular-lnahlh-hw2d3b

Answer 2

We must say how many rows there, but some rows has same id and we will sort and merge td if them used same id. But for your data, it is say some rows there, and descriptions is array and split able. For this way JS can't know how many <tr> should be there.

2 ways for you: 1- Format your data, keep one description each row, same as the example data in second href, [{id, name, weight, countdescriptions, description},...], and use [attr.rowspan]='data.countdescriptions' instead [attr.rowspan]='getRowSpan(data.id)'. 2- Update content format, like <ul><li *ngFor... in the description <td>, and remove the [attr.rowspan] attribute.

Answer 3

Well, it seems that material table has no api documentation for it, I could not find any trick to do this too, But we can twick the our data to support this, as per your second example we can reform the data to new json and we can get our expected result.

Step 1 :

    const originalData = [
      { id: 1, name: 'Hydrogen', weight: 1.0079, descriptions: ['row1', 'row2'] },
      { id: 2, name: 'Helium', weight: 4.0026, descriptions: ['row1', 'row2', 'row3'] },
      { id: 3, name: 'Lithium', weight: 6.941, descriptions: ['row1', 'row2'] },
      { id: 4, name: 'Beryllium', weight: 9.0122, descriptions: ['row1', 'row2', 'row3'] },
      { id: 5, name: 'Boron', weight: 10.811, descriptions: ['row1'] },
      { id: 6, name: 'Carbon', weight: 12.0107, descriptions: ['row1', 'row2', 'row3'] },
      { id: 7, name: 'Nitrogen', weight: 14.0067, descriptions: ['row1'] },
      { id: 8, name: 'Oxygen', weight: 15.9994, descriptions: ['row1'] },
      { id: 9, name: 'Fluorine', weight: 18.9984, descriptions: ['row1', 'row2', 'row3'] },
      { id: 10, name: 'Neon', weight: 20.1797, descriptions: ['row1', 'row2', 'row3'] },
    ]; //original data

    const DATA = originalData.reduce((current, next) => {
      next.descriptions.forEach(b => {
        current.push({ id: next.id, name: next.name, weight: next.weight, descriptions: b })
      });
      return current;
    }, []);//iterating over each one and adding as the description 
    console.log(DATA)

    const ELEMENT_DATA: PeriodicElement[] = DATA; //adding data to the element data

Step 2

this will be as it is as your second stackblitz link.

 getRowSpan(col, index) {    
    return this.spans[index] && this.spans[index][col];
  }

Step 3

as it is as per your second link

  constructor() {
    this.cacheSpan('Priority', d => d.id);
    this.cacheSpan('Name', d => d.name);
    this.cacheSpan('Weight', d => d.weight);
  }

  /**
   * Evaluated and store an evaluation of the rowspan for each row.
   * The key determines the column it affects, and the accessor determines the
   * value that should be checked for spanning.
   */
  cacheSpan(key, accessor) {
    for (let i = 0; i < DATA.length;) {
      let currentValue = accessor(DATA[i]);
      let count = 1;

      // Iterate through the remaining rows to see how many match
      // the current value as retrieved through the accessor.
      for (let j = i + 1; j < DATA.length; j++) {
        if (currentValue != accessor(DATA[j])) {
          break;
        }

        count++;
      }

      if (!this.spans[i]) {
        this.spans[i] = {};
      }

      // Store the number of similar values that were found (the span)
      // and skip i to the next unique row.
      this.spans[i][key] = count;
      i += count;
    }
  }

Step 4

Using index to pass down to rowspan and hiding the rows where it doesn't needed

    <ng-container matColumnDef="id">
        <th mat-header-cell *matHeaderCellDef> Priority </th>
        <td mat-cell *matCellDef="let data;let i = dataIndex" [attr.rowspan]="getRowSpan('Priority',i)" [style.display]="getRowSpan('Priority', i) ? '' : 'none'">
         {{ data.id }} </td>
    </ng-container>

    <ng-container matColumnDef="name">
        <th mat-header-cell *matHeaderCellDef> Name </th>
        <td mat-cell *matCellDef="let data;let i = dataIndex" [attr.rowspan]="getRowSpan('Name',i)" [style.display]="getRowSpan('Name', i) ? '' : 'none'">
         {{ data.name }} </td>
    </ng-container>

    <ng-container matColumnDef="weight">
        <th mat-header-cell *matHeaderCellDef> Weight </th>
        <td mat-cell *matCellDef="let data;let i = dataIndex" [attr.rowspan]="getRowSpan('Weight',i)" [style.display]="getRowSpan('Weight', i) ? '' : 'none'">
         {{ data.weight }} </td>
    </ng-container>

    <ng-container matColumnDef="descriptions">
        <th mat-header-cell *matHeaderCellDef> Descriptions </th>
        <td mat-cell *matCellDef="let data"> {{ data.descriptions }} </td>
    </ng-container>

    <tr mat-header-row *matHeaderRowDef="displayedColumns"></tr>
    <tr mat-row *matRowDef="let row; columns: displayedColumns;"></tr> 


</table>

Here is the demo