0.6.0 globs
i confused
-
correct?
- https://github.com/terrastruct/d2/issues/1487#issuecomment-1657459438
-
(lady* -> barbie)[*].style.stroke: pink https://d2lang.com/tour/globs/#glob-connections
- what is '[ ]' on d2? ( [0]? [1]? ... [*]?)
-
** is recursive?
- "ex. zone-A.** -> load balancer"
-
good to see you again @bo-ku-ra !
- Yes, correct for now, but it's temporary. Later this week we will change globs to apply to objects made after it. But right now it only applies to objects already made.
- It's the connection index. https://play.d2lang.com/?script=qlDQtVOo5IJSXBoQhma0Qawe2EdWCpkKibkKhggZQ1QZIy5AAAAA__8%3D&
- We have a note in the docs that the recursive doesn't get containers when used on connections. It's for practical reasons -- we couldn't think of a case where someone would want the behavior to also point containers to something.
3.We have a note in the docs that the recursive doesn't get containers when used on connections. It's for practical reasons -- we couldn't think of a case where someone would want the behavior to also point containers to something.
'x', 'y', 'z' and 'bar' are same layer. https://play.d2lang.com/?l=&script=ZJHNTsMwEITvfoqhOQQsdVPlwMFIcOUdEEJ23DgWSTZyzE9J8-4oTisqevJ4P3nGq6mZFSYBfAvgIACjwzo4yc0GU_qmgu-02z9g2VmhiXEYVVHYstW9o4q7wneusOWbC3pofEXjp5uTT8o4djZdgAzP_iS3sCV2dE87-BEVc3uTyFEAy9sfMQtRMxsd0klSYvuI8yRbFBkdqGb-Awr9R9teUKPDFb09wwvHuxf5umJksNznEV8c3p_W7H8uA_s-IjJi40eYA3Ipc0ypG1rbUgh7O4vfAAAA__8%3D&
- i think '**' should apply to all.
- if i don't need them, i can delete them with 'null'.
this is simple. i can easily understand it.
if svg is output by the auto-output program then i can not imagine the correct diagrams. (you can find out all the nests correctly, right?)
@alixander https://github.com/terrastruct/d2/issues/1511#issuecomment-1657747186
We have a note in the docs that the recursive doesn't get containers when used on connections. It's for practical reasons -- we couldn't think of a case where someone would want the behavior to also point containers to something.
the situation has changed since v0.6.2 (the ELK layout engine will now route SQL diagrams to their exact columns.)
expect
- a.g.h.i -> z
- a.c.e -> z
result
vars: {
d2-config: {
layout-engine: elk
}
}
a.b
a.c.d
a.c.e.f
a.g.h.i: {
foobar: {
shape: sql_table
integer: int {constraint: [primary_key; unique]}
text: char {constraint: foreign_key}
json: jsonb {constraint: unique}
}
}
** -> z
https://play.d2lang.com/?script=VE9BbsMwDLsbyB94LhAfdvSAfWQYCiVVHK2Z3MrOsKzI3wevW9tcKIoiCeiTLAdcGgccnto-6SDxbwcmWtJcWtYoygE8Hau-Nm5tXOPIdxV6f7gO9kMl0Y9ebhXAkFJH9iAAeaQTB-TztC_UTXy_iBaObKESXPqkuRiJloDXk8kH2bI_8vKMWeU889t6Txb-KgH9SLbNDclYotbYg_s9Jw2_2G3t1-Kb8__X3Q7tC77dTwAAAP__&
i think this specification is difficult to imagine and use.