[0.11.2] Plain text parsing gives strange results #781

New Issue

2017-08-17T09:01:49Z

the-spyke commented

2017-08-17 09:01:49 +00:00

(Migrated from github.com)

Version [0.11.2] has plain text parsing enabled by default.

The default value for raw setting isn't specified in Parser Options table
Its behavior is strange

Source

const workbook = XLSX.read(base64String, {
    type: "base64",
    cellHTML: false,
    cellText: false,
});

Input CSV

sku,day,store,color,sales
sku 24,20120101,store-1,blue,$44.00

Result workbook object

{
    "SheetNames": [
        "Sheet1"
    ],
    "Sheets": {
        "Sheet1": {
            "A1": {
                "t": "s",
                "v": "sku"
            },
            "B1": {
                "t": "s",
                "v": "day"
            },
            "C1": {
                "t": "s",
                "v": "store"
            },
            "D1": {
                "t": "s",
                "v": "color"
            },
            "E1": {
                "t": "s",
                "v": "sales"
            },
            "A2": {
                "t": "s",
                "v": "sku 24"
            },
            "B2": {
                "t": "n",
                "w": "20120101",
                "v": 20120101
            },
            "C2": {
                "z": "m/d/yy",
                "t": "n",
                "v": 36892,
                "w": "1/1/01"
            },
            "D2": {
                "t": "s",
                "v": "blue"
            },
            "E2": {
                "t": "n",
                "w": "$44.00",
                "v": 44
            },
            "!ref": "A1:E2"
        }
    }
}

What's wrong

store-1 is parsed as 36892
It's cell has z property even if cellNF is false by default
Some cells have w property even if cellText is false in supplied options object

Conclusions

Not sure if raw should be false by default, but I'm fine with that
Please add the default value for it into the Parser Options table
Need to update plain text parsing logic?

Version [0.11.2] has plain text parsing enabled by default. 1. The default value for `raw` setting isn't specified in `Parser Options` table 2. Its behavior is strange ## Source ``` const workbook = XLSX.read(base64String, { type: "base64", cellHTML: false, cellText: false, }); ``` ## Input CSV ``` sku,day,store,color,sales sku 24,20120101,store-1,blue,$44.00 ``` ## Result workbook object ``` { "SheetNames": [ "Sheet1" ], "Sheets": { "Sheet1": { "A1": { "t": "s", "v": "sku" }, "B1": { "t": "s", "v": "day" }, "C1": { "t": "s", "v": "store" }, "D1": { "t": "s", "v": "color" }, "E1": { "t": "s", "v": "sales" }, "A2": { "t": "s", "v": "sku 24" }, "B2": { "t": "n", "w": "20120101", "v": 20120101 }, "C2": { "z": "m/d/yy", "t": "n", "v": 36892, "w": "1/1/01" }, "D2": { "t": "s", "v": "blue" }, "E2": { "t": "n", "w": "$44.00", "v": 44 }, "!ref": "A1:E2" } } } ``` ## What's wrong - `store-1` is parsed as `36892` - It's cell has `z` property even if `cellNF` is `false` by default - Some cells have `w` property even if `cellText` is `false` in supplied options object ## Conclusions - Not sure if `raw` should be `false` by default, but I'm fine with that - Please add the default value for it into the `Parser Options` table - Need to update plain text parsing logic?

SheetJSDev commented

2017-08-18 18:19:27 +00:00

(Migrated from github.com)

README updated to mention that raw defaults to false
CSV parsing honors cellText and cellNF

Most of the ancillary options were initially added to XLSX or XLS but haven't been propagated to the other formats.

- [README](https://github.com/sheetjs/js-xlsx#parsing-options) updated to mention that `raw` defaults to `false` - CSV parsing honors `cellText` and `cellNF` Most of the ancillary options were initially added to XLSX or XLS but haven't been propagated to the other formats.

👍 1

Sign in to join this conversation.