有时我会在触发断点时得到这个.看起来堆栈帧没有被保存,所以我无法退回调用堆栈 – 真正的痛苦.请参阅下面的示例
--> #0 BatchProcess.add_failure_record(row_id#Fixnum,test#Struct::Test,message#String,...)
at line server/processes/batch.rb:309
Warning: saved frames may be incomplete; compare with caller(0).
(rdb:1) pp caller
["./server/processes/batch.rb:309:in `run_tests'","./server/processes/common/generic_process.rb:219:in `each'","./server/processes/common/generic_process.rb:219:in `run_tests'","./server/processes/common/generic_process.rb:271:in `run_plan'","./server/processes/common/corrections.rb:19:in `each_with_index'","./server/processes/common/generic_process.rb:266:in `each'","./server/processes/common/generic_process.rb:266:in `each_with_index'","./server/processes/common/generic_process.rb:266:in `run_plan'","./server/processes/batch.rb:202:in `run_engine'","/usr/lib/ruby/1.8/benchmark.rb:293:in `measure'","./server/processes/batch.rb:201:in `run_engine'","./server/processes/common/generic_process.rb:88:in `run_dataset'","./server/processes/batch.rb:210:in `run_dataset'","./server/processes/batch.rb:209:in `run_dataset'","./server/processes/common/generic_process.rb:159:in `run'","./server/processes/common/generic_process.rb:158:in `each'","./server/processes/common/generic_process.rb:158:in `run'","./server/processes/batch.rb:350:in `run'","./server/processes/batch.rb:349:in `run'","server/processes/test_runs/run_tests.rb:55:in `run_one_process'","server/processes/test_runs/run_tests.rb:81"]
关于如何阻止这种情况发生的任何想法？
解决方法
这意味着Ruby的caller()函数报告的调用堆栈与调试器记录的调用堆栈不匹配.
如果在调用堆栈中有多个帧时激活了调试器跟踪,则会发生这种情况.它也可能由于ruby-debug中的错误导致无法正确跟踪.一个“补救措施”是将它放在主文件中文件的最开头：
require’ruby-debug’; Debugger.start
然而,这方面的缺点是从程序开始就添加了一些额外的开销.
在Ruby代码中的某处,您必须告诉Ruby开始跟踪调用和返回.通常,这是通过Debugger.start和Debugger.stop或Debugger.start {}的组合完成的.如果您从框架运行,这些命令将在某处发布,因此在正确的位置发布的责任在于添加此代码的程序员.
最后,我将更改消息更加明确：
Warning: saved frames may be incomplete; compare debugger backtrace (bt) with Ruby caller(0).