Without seeing actual code, I find it difficult to understand your explanation enough to come up with any guess. But it seems you have 3 things: leo code, your comparison code, and your unit test code. Since this problem started happening, at least one of those things had to have changed right? I see you are running a git commit from a month ago, so maybe your leo code hasn't changed since the problem appeared and so it must be your unit test or program code? Do you have a history (in git or something) of your code? When it is a problem which is fairly easy to reliably reproduce then git bisect is very handy.
It isn't clear from your description what your 2nd subprocess.call is executing. Whatever it is, maybe you can have extra tracing or set a debugger breakpoint whenever newHeadline is seen and that way you can get more information. On Thu, Sep 19, 2019 at 6:02 PM SegundoBob <[email protected]> wrote: > > *Leo-Editor Version:* > Leo 6.1-devel, devel branch, build f5986d8016 > 2019-08-09 10:23:41 -0500 > Python 3.6.8, PyQt version 5.11.1 > > *Bug Description:* > > This bug may be in my code or in Leo-Editor code. I have been seeing > occurrences for at least two weeks. > > The bug occurs when I run my unit tests for my program that compares two > Leo-Editor files. I have 41 unit tests. The bug occurs about once every > 30 tests. > > My comparison program uses LeoBridge to access Leo-Editor files. File > "Content" as accessed by vnode (not position). File "structure" is > analyzed by position (path). That is, my program gets the headline and > body from the vnode (vn.h, bn.b). When my comparison program compares the > unit test File A and File B, the bug never occurs. Using LeoBridge my > comparision program produces a difference file (File D). The unit tests > use subprocess.call() to execute my comparision program, so the process > comparing File A and File B and producing File D ends. Then the unit tests > use subprocess.call() to compare the actual File D with the expected File > D. This is when the bug occurs. > > The bug is always that a headline in File D, is seen to be "NewHeadline." > But when I look at File D, there is no "NewHeadline" and File D is exactly > as expected. > > There are never two or more bug occurrences in comparing one File D. The > bug never occurs on a node body. There is a slight tendency for bug > occurrences to cluster. That is, sometimes it occurs on two tests in a row > or even three tests in row. > > *Help:* > > Does anyone have any ideas about how this bug could occur? Does anyone > have any ideas about how to chase this bug? > > -- > You received this message because you are subscribed to the Google Groups > "leo-editor" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > To view this discussion on the web visit > https://groups.google.com/d/msgid/leo-editor/fd49a5e7-6694-4bf6-9b22-88c517ef38ef%40googlegroups.com > <https://groups.google.com/d/msgid/leo-editor/fd49a5e7-6694-4bf6-9b22-88c517ef38ef%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- You received this message because you are subscribed to the Google Groups "leo-editor" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/leo-editor/CAO5X8CzxvJ32vuYGr%3DfqXh0quQi22xxZYQuUQX6AyeH7p7SR2Q%40mail.gmail.com.
