Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

v10.10.1 Operator: contains regression #675

Closed
VictoriaHunsaker opened this issue Mar 4, 2024 · 2 comments · May be fixed by bossazizan/docs#4, OATEF4313/docs#1 or Playgirlkaybraz11/docs#6
Closed

v10.10.1 Operator: contains regression #675

VictoriaHunsaker opened this issue Mar 4, 2024 · 2 comments · May be fixed by bossazizan/docs#4, OATEF4313/docs#1 or Playgirlkaybraz11/docs#6
Labels

Comments

@VictoriaHunsaker
Copy link

Hello, I think the latest release (10.10.1) has introduced a bug in the contains operator.

Given this template:

{% assign str_var = "xyz" %}
{% assign same_str_var = "xyz" %}
{% assign diff_str_var = "abc" %}

{{ str_var contains "xy" }}
{{ str_var contains same_str_var }}
{{ str_var contains diff_str_var }}
{{ "xyz" contains str_var }}

In 10.10.0, the result is

true
true
false
true

However, in 10.10.1, the result is

false
false
false
true
@harttle
Copy link
Owner

harttle commented Mar 9, 2024

Hi @VictoriaHunsaker , thanks for the report. I'm not able to repro the results, the latest version here yields results same as In 10.10.0 you mentioned. But I can see there's a change in contains between v10.10.0 and v10.10.1: 1937aa1

Could you provide more information:

  • How did you introduce LiquidJS, via npm or via external script? For the latter, could you provide the link of the LiquidJS javascript file you're using?
  • The environment you reproduced this problem. Could you provide Node.js version if in Node.js, or your UserAgent (window.navigator.userAgent) if in browser?

@harttle harttle added the bug label Mar 9, 2024
@benweissmann
Copy link

Hey, i'm working on the same project as @VictoriaHunsaker and dug into this a bit -- our issue is that we're passing in a wrapper value around user-provided string, to throw an explicit error if someone tries to use a non-numeric value for an operator like > or < or a filter like plus (rather than the default behavior of coercing the string to NaN). So the new check for isString(l) is introducing this regression, because we're providing an object with a indexOf() method rather than a String object:

class StrictStringForLiquid {
  private name: string;
  private value: string;

  constructor(name: string, value: string) {
    this.name = name;
    this.value = value;
  }

  toString() {
    return this.value;
  }

  valueOf() {
    const num = Number(this.value);
    if (isNaN(num)) {
      throw new Error(
        `expected ${this.name} to be a number, but it was: "${this.value}"`
      );
    }

    return num;
  }

  equals(other: unknown) {
    return this.value === String(other);
  }

  gt(other: unknown) {
    return this.valueOf() > Number(other);
  }

  geq(other: unknown) {
    return this.valueOf() >= Number(other);
  }

  lt(other: unknown) {
    return this.valueOf() < Number(other);
  }

  leq(other: unknown) {
    return this.valueOf() <= Number(other);
  }

  indexOf(other: unknown) {
    return this.value.indexOf(String(other));
  }
}

Note that Drops don't work for this use-case; the valueOf method on a Drop gets called regardless of context, whereas with a custom non-drop object like this, valueOf gets called in contexts where Liquid expects a number, and toString gets called in contexts where it expects a string.

I think you could preserve backwards-compatibility by replacing the isString(l) conditional with l && isFunction(l.indexOf)?

harttle added a commit that referenced this issue Mar 21, 2024
* fix: `contains` regression on string-like objects, #675

* chore: fix build docs on macos
github-actions bot pushed a commit that referenced this issue Mar 21, 2024
## [10.10.2](v10.10.1...v10.10.2) (2024-03-21)

### Bug Fixes

* contains regression ([#677](#677)) ([05223c4](05223c4)), closes [#675](#675)
@harttle harttle closed this as completed Apr 28, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
3 participants