LANG-1695: NumberUtils::isParseable does not recognise numbers ending in dot #1071

orionlibs · 2023-06-29T21:25:54Z

LANG-1695 allowed for a number (positive or negative) that has only one decimal point and it is at the end of the input, is treated as a valid number

…al point and it is at the end of the input, is treated as a valid number

codecov-commenter · 2023-06-29T21:41:23Z

Codecov Report

Merging #1071 (78695e1) into master (4756eb0) will decrease coverage by 0.01%.
Report is 99 commits behind head on master.
The diff coverage is 0.00%.

@@             Coverage Diff              @@
##             master    #1071      +/-   ##
============================================
- Coverage     92.13%   92.13%   -0.01%     
  Complexity     7504     7504              
============================================
  Files           195      195              
  Lines         15722    15720       -2     
  Branches       2897     2896       -1     
============================================
- Hits          14486    14483       -3     
  Misses          665      665              
- Partials        571      572       +1

Files Changed	Coverage Δ
...ava/org/apache/commons/lang3/math/NumberUtils.java	`94.77% <0.00%> (-0.25%)`	⬇️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

garydgregory · 2023-07-02T14:55:20Z

@orionlibs
Thank you for the PR.

Would you please reply in the Jira ticket to the comment https://issues.apache.org/jira/browse/LANG-1695?focusedCommentId=17709278&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-17709278

We really need to make sure we are not creating trouble for ourselves ;-)

orionlibs · 2023-07-02T15:11:37Z

@garydgregory I am new to open source contribution and a week ago I started with Apache. I applied for a JIRA account, but I was denied one (with no reason given). So, I cannot login to JIRA. Here is the email from apache:
We regret to inform you that, upon reviewing your request for a new Jira account connected with The Apache Software Foundation,
the commons project has chosen to deny the request. We therefore will not create the Jira account.

The following reason was given:
No reason given.

If you wish to appeal this decision, you may do so either directly with the
commons project, or by contacting the ASF Infrastructure team at:
users@infra.apache.org

orionlibs · 2023-07-02T15:12:58Z

@garydgregory also there are many many open tickets in JIRA for many projects that have actually been completed, but they are still open, with no assignees or comments or anything.

garydgregory · 2023-07-02T16:19:18Z

@garydgregory I am new to open source contribution and a week ago I started with Apache. I applied for a JIRA account, but I was denied one (with no reason given). So, I cannot login to JIRA. Here is the email from apache: We regret to inform you that, upon reviewing your request for a new Jira account connected with The Apache Software Foundation, the commons project has chosen to deny the request. We therefore will not create the Jira account.

The following reason was given: No reason given.

If you wish to appeal this decision, you may do so either directly with the commons project, or by contacting the ASF Infrastructure team at: users@infra.apache.org

What user name and real name did you use in the application? I'll see if I can find the application and rejection in the archives.

orionlibs · 2023-07-02T16:27:02Z

What user name and real name did you use in the application? I'll see if I can find the application and rejection in the archives.

@garydgregory I used this page
https://selfserve.apache.org/jira-account.html
and the info I entered was:
email address = efthymiou.dimitrios1@gmail.com
username = dimitrios.efthymiou
real public name = Dimitrios Efthymiou
tell us the intent for this account = I don't need the account now, but I don't know when I will. It is like when you join a company. They give you access to some basic services. (This was probably the reason, but I didn't know then that I needed access to JIRA in order to update the tickets I worked on. I worked on a few tickets that were unassigned and did not have PR links in them. I know that now. With a JIRA account I can assign myself to the tickets I work on and I can add my PR links in there. I guess that was the reason they rejected me. Maybe I can register once again, but with a better reason this time?)

garydgregory · 2023-07-03T00:40:30Z

@garydgregory I am new to open source contribution and a week ago I started with Apache. I applied for a JIRA account, but I was denied one (with no reason given). So, I cannot login to JIRA. Here is the email from apache: We regret to inform you that, upon reviewing your request for a new Jira account connected with The Apache Software Foundation, the commons project has chosen to deny the request. We therefore will not create the Jira account.

The following reason was given: No reason given.

If you wish to appeal this decision, you may do so either directly with the commons project, or by contacting the ASF Infrastructure team at: users@infra.apache.org

@orionlibs
Please apply again for a Jira account, I'll watch for it.

orionlibs · 2023-07-03T09:17:00Z

thank you Gary. I just registered again for JIRA with the email = ***@***.*** username = dimitrios.efthymiou

…

On Mon, 3 Jul 2023 at 01:40, Gary Gregory ***@***.***> wrote: @garydgregory <https://github.com/garydgregory> I am new to open source contribution and a week ago I started with Apache. I applied for a JIRA account, but I was denied one (with no reason given). So, I cannot login to JIRA. Here is the email from apache: We regret to inform you that, upon reviewing your request for a new Jira account connected with The Apache Software Foundation, the commons project has chosen to deny the request. We therefore will not create the Jira account. The following reason was given: No reason given. If you wish to appeal this decision, you may do so either directly with the commons project, or by contacting the ASF Infrastructure team at: ***@***.*** @orionlibs <https://github.com/orionlibs> Please apply again for a Jira account, I'll watch for it. — Reply to this email directly, view it on GitHub <#1071 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABGQOMN5ITPI66OJCP5XUBTXOIIIVANCNFSM6AAAAAAZZBH2WU> . You are receiving this because you were mentioned.Message ID: ***@***.***>

elharo · 2023-07-12T13:01:43Z

src/test/java/org/apache/commons/lang3/math/NumberUtilsTest.java

@@ -918,7 +918,8 @@ public void testIsParsable() {
        assertFalse(NumberUtils.isParsable("pendro"));
        assertFalse(NumberUtils.isParsable("64, 2"));
        assertFalse(NumberUtils.isParsable("64.2.2"));
-        assertFalse(NumberUtils.isParsable("64."));
+        assertTrue(NumberUtils.isParsable("64."));


please add a test that 64.. is false

@elharo sorry. My understanding of the ticket is that 64. is true i.e. parseable and valid

is valid but 64.. (two periods) is not

you are right. I added a test for it

…to LANG-1695-NumberUtils-isParseable-does-not-recognise-values-such-as-2-dot

garydgregory · 2023-07-28T09:41:40Z

src/main/java/org/apache/commons/lang3/math/NumberUtils.java

@@ -1776,7 +1773,7 @@ private static boolean withDecimalsParsing(final String str, final int beginIdx)
            if (decimalPoints > 1) {
                return false;
            }
-            if (!isDecimalPoint && !Character.isDigit(str.charAt(i))) {
+            if (!isDecimalPoint && !Character.isDigit(str.charAt(i)) && !(i == str.length() - 1 && str.charAt(i) == '.')) {


What happens in countries that flip the meaning of the comma and period?

perhaps there could be another method that takes a Locale object. Otherwise, I don't think the current parser should concern itself with it

Per current Javadoc "Parsable numbers include those Strings understood by Integer#parseInt(String), Long#parseLong(String), Float#parseFloat(String)} or Double#parseDouble(String)."

If those methods don't worry about countries that flip the meaning of the comma and period, we shouldn't do it here. That would be a breaking behavior change.

As to another method that uses Locales, that's in java.util so outside the advertised scope for commons-lang.

elharo · 2023-07-29T11:32:11Z

src/main/java/org/apache/commons/lang3/math/NumberUtils.java

@@ -1776,7 +1773,7 @@ private static boolean withDecimalsParsing(final String str, final int beginIdx)
            if (decimalPoints > 1) {
                return false;
            }
-            if (!isDecimalPoint && !Character.isDigit(str.charAt(i))) {
+            if (!isDecimalPoint && !Character.isDigit(str.charAt(i)) && !(i == str.length() - 1 && str.charAt(i) == '.')) {


Per current Javadoc "Parsable numbers include those Strings understood by Integer#parseInt(String), Long#parseLong(String), Float#parseFloat(String)} or Double#parseDouble(String)."

If those methods don't worry about countries that flip the meaning of the comma and period, we shouldn't do it here. That would be a breaking behavior change.

As to another method that uses Locales, that's in java.util so outside the advertised scope for commons-lang.

garydgregory · 2023-07-29T19:11:50Z

To make things more complicated, in the French locale a non-breaking space can be used to separate some items. I can't recall the details now but it's yet another thing to lookout for.

elharo · 2023-07-29T20:10:02Z

The more I think about this, the more I'm convinced we should not account for localization. This method is not for the purpose of determining whether a string is plausibly a number. Its purpose is to return true or false depending on whether the string can be parsed by one of the java.lang.Number classes. Those classes don't consider localization, so this one shouldn't either.

garydgregory · 2023-07-29T21:26:11Z

@elharo
I agree with you on the purpose of the API.
I'd like to revisit the implementation though: Why are we half-baked-with-bugs parsing the input instead of just try-catching what we claim is input that will or not throw an exception when we a user actually calls on the number APIs?

elharo · 2023-08-20T10:56:29Z

One purpose of this method is precisely to avoid using exceptions for flow control.

orionlibs · 2023-12-03T13:35:11Z

this PR is still open after half a year

--allowed for a number (positive or negative) that has only one decim…

75393a3

…al point and it is at the end of the input, is treated as a valid number

elharo suggested changes Jul 12, 2023

View reviewed changes

orionlibs changed the title ~~LANG-1695~~ LANG-1695: NumberUtils::isParseable does not recognise numbers ending in dot Jul 17, 2023

orionlibs added 2 commits July 28, 2023 08:16

Merge branch 'master' of https://github.com/orionlibs/commons-lang in…

95d410d

…to LANG-1695-NumberUtils-isParseable-does-not-recognise-values-such-as-2-dot

LANG-1695: added test for special case

78695e1

garydgregory reviewed Jul 28, 2023

View reviewed changes

elharo approved these changes Jul 29, 2023

View reviewed changes

orionlibs closed this by deleting the head repository Apr 14, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

LANG-1695: NumberUtils::isParseable does not recognise numbers ending in dot #1071

LANG-1695: NumberUtils::isParseable does not recognise numbers ending in dot #1071

orionlibs commented Jun 29, 2023 •

edited

codecov-commenter commented Jun 29, 2023 •

edited

garydgregory commented Jul 2, 2023 •

edited

orionlibs commented Jul 2, 2023 •

edited

orionlibs commented Jul 2, 2023

garydgregory commented Jul 2, 2023

orionlibs commented Jul 2, 2023 •

edited

garydgregory commented Jul 3, 2023

orionlibs commented Jul 3, 2023 via email

elharo Jul 12, 2023

orionlibs Jul 17, 2023 •

edited

elharo Jul 27, 2023

orionlibs Jul 28, 2023

garydgregory Jul 28, 2023

orionlibs Jul 28, 2023

elharo Jul 29, 2023

elharo Jul 29, 2023

garydgregory commented Jul 29, 2023

elharo commented Jul 29, 2023

garydgregory commented Jul 29, 2023

elharo commented Aug 20, 2023

orionlibs commented Dec 3, 2023

LANG-1695: NumberUtils::isParseable does not recognise numbers ending in dot #1071

LANG-1695: NumberUtils::isParseable does not recognise numbers ending in dot #1071

Conversation

orionlibs commented Jun 29, 2023 • edited

codecov-commenter commented Jun 29, 2023 • edited

Codecov Report

garydgregory commented Jul 2, 2023 • edited

orionlibs commented Jul 2, 2023 • edited

orionlibs commented Jul 2, 2023

garydgregory commented Jul 2, 2023

orionlibs commented Jul 2, 2023 • edited

garydgregory commented Jul 3, 2023

orionlibs commented Jul 3, 2023 via email

elharo Jul 12, 2023

Choose a reason for hiding this comment

orionlibs Jul 17, 2023 • edited

Choose a reason for hiding this comment

elharo Jul 27, 2023

Choose a reason for hiding this comment

orionlibs Jul 28, 2023

Choose a reason for hiding this comment

garydgregory Jul 28, 2023

Choose a reason for hiding this comment

orionlibs Jul 28, 2023

Choose a reason for hiding this comment

elharo Jul 29, 2023

Choose a reason for hiding this comment

elharo Jul 29, 2023

Choose a reason for hiding this comment

garydgregory commented Jul 29, 2023

elharo commented Jul 29, 2023

garydgregory commented Jul 29, 2023

elharo commented Aug 20, 2023

orionlibs commented Dec 3, 2023

orionlibs commented Jun 29, 2023 •

edited

codecov-commenter commented Jun 29, 2023 •

edited

garydgregory commented Jul 2, 2023 •

edited

orionlibs commented Jul 2, 2023 •

edited

orionlibs commented Jul 2, 2023 •

edited

orionlibs Jul 17, 2023 •

edited