make: *** No rule to make target ... package-info.java', needed by 'biostar140111'. Stop.

[mmokrejs]

Hi,
it seems the Makefile does not ensure that a make exits upon error so nobody realized this so far? Below is a current HEAD checkout:

$ make all
...
mkdir -p /var/tmp/portage/sci-biology/jvarkit-9999/work/jvarkit-9999/src/main/generated-sources/java
xjc -d /var/tmp/portage/sci-biology/jvarkit-9999/work/jvarkit-9999/src/main/generated-sources/java  -p gov.nih.nlm.ncbi.dbsnp  "http://ftp.ncbi.nlm.nih.gov/snp/specs/docsum_current.xsd"
^M100 7628k    0 7628k    0     0   466k      0 --:--:--  0:00:16 --:--:--  354kparsing a schema...
^M100 7835k    0 7835k    0     0   451k      0 --:--:--  0:00:17 --:--:--  246k^M100 8033k    0 8033k    0     0   438k      0 --:--:--  0:00:18 --:--:--  200kcompiling a schema...
gov/nih/nlm/ncbi/dbsnp/Assay.java
gov/nih/nlm/ncbi/dbsnp/Assembly.java
gov/nih/nlm/ncbi/dbsnp/BaseURL.java
gov/nih/nlm/ncbi/dbsnp/Component.java
gov/nih/nlm/ncbi/dbsnp/ExchangeSet.java
gov/nih/nlm/ncbi/dbsnp/FxnSet.java
gov/nih/nlm/ncbi/dbsnp/MapLoc.java
gov/nih/nlm/ncbi/dbsnp/ObjectFactory.java
gov/nih/nlm/ncbi/dbsnp/PrimarySequence.java
gov/nih/nlm/ncbi/dbsnp/Rs.java
gov/nih/nlm/ncbi/dbsnp/RsLinkout.java
gov/nih/nlm/ncbi/dbsnp/RsStruct.java
gov/nih/nlm/ncbi/dbsnp/Ss.java
gov/nih/nlm/ncbi/dbsnp/package-info.java
mkdir -p /var/tmp/portage/sci-biology/jvarkit-9999/work/jvarkit-9999/src/main/generated-sources/java
make: *** No rule to make target '/var/tmp/portage/sci-biology/jvarkit-9999/work/jvarkit-9999/src/main/generated-sources/java/gov/nih/nlm/ncbi/dbsnp/gt/package-info.java', needed by 'biostar140111'.  Stop.
make: *** Waiting for unfinished jobs....

[lindenb]

this is strange because the current head is using

the following hash: 8dd4559dced3a243dde094415a9974c03aceb3e3

and not what I'm seeing in you trace: 9999 . How did you clone that project ?

furthermore, 'make all' works in travis: https://travis-ci.org/lindenb/jvarkit#L8149

[lindenb]

Anyway, the Makefile rules is missing, I think that is what the message says.

I don't see any rule missing.

make -nB all produces no error.

furthermore, I suspect you're using a non-GNU version of make

(...)
 *   emake failed

And you can always restructure the makefile so that there are always these separate targets
'fetch', 'compile', 'install', 'all'

this is not my current interest, but feel free to submit a PR.

[lindenb]

So what does the 'No rule to make target' mean?

I know it's not your final aim, but I would suggest you use a standard version of make, not your wrapper, and test if you get the very same message. If not, then I would say that it comes from your wrapper emake.

[lindenb]

OK, I re-tried with calling make directly. I also forced one thread ... I got a different error.

/var/tmp/portage/sci-biology/jvarkit-9999/work/jvarkit-9999/src/main/java/com/github/lindenb/jvarkit/tools/misc/VcfToHilbert.java:162: error: unmappable character
 for encoding ASCII
        // Make A-shaped (really "???" shaped) curve at this scale:
                                    ^

Ah, I agree that one. There is an unicode character in my java file that was not detected by my tools.

file ./main/java/com/github/lindenb/jvarkit/tools/misc/VcfToHilbert.java
./main/java/com/github/lindenb/jvarkit/tools/misc/VcfToHilbert.java: UTF-8 Unicode C program text

it looks like, it's the only file with this problem:

$ find -type f -exec file '{}' ';' | grep Unicode
./main/java/com/github/lindenb/jvarkit/tools/misc/VcfToHilbert.java: UTF-8 Unicode C program text

fixed in c072292

$ file ./main/java/com/github/lindenb/jvarkit/tools/misc/VcfToHilbert.java./main/java/com/github/lindenb/jvarkit/tools/misc/VcfToHilbert.java: ASCII C program text

thanks

[lindenb]

fixed (via email)