Check for ast.Attributes when finding occurrences in f-strings

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!

[sandratsy]

Description

Given the following piece of code:

defxyz(self):
print(f'{self.abc}')
returnself.abc

If I try to Rename attribute/property abc, the abc in the return statement is correctly renamed, while the abc in the f-string is not. This is because _search_in_f_string() only looks for ast.Names, while self.abc is an ast.Attribute. I have updated _search_in_f_string() to look for ast.Attributes as well.

(Ofc, this is not fool-proof, since devs can put whatever they like in f-strings, but I suppose devs are more likely to put Names and Attributes in f-strings than entire function calls.)

The node.col_offset for an ast.Attribute points to the start of self.abc rather than abc, however. To get around this, I've taken the liberty to change _search_in_f_string() to return an offset instead of the node. node.end_col_offset - len(self.name) will correctly return the index of abc.

I used node.end_col_offset - len(self.name) instead of node.col_offset + f_string.index(self.name) in case the attribute is named abc.abc. We want the offset of the 2nd abc, not the 1st.

Checklist (delete if not relevant):

• I have updated CHANGELOG.md

[sandratsy]

Hello, any updates on this? Is there anything else I should do?

[lieryan]

Thanks for the contribution @sandratsy, this is great. I've added some test cases and type hints but the fix looks sensible.

[lieryan]

@all-contributors please add @sandratsy for code

[allcontributors]

@lieryan

I've put up a pull request to add @sandratsy! 🎉

[sandratsy]

Thanks v much!