The fix is to append a newline to the source being evaluated.
Without this patch, the following commands **print no output, no errors**.
```
(lldb) script if "foo" in lldb.frame.name: print(lldb.thread)
(lldb) script for f in lldb.thread: print(f.name)
```
The issue is with `code.InteractiveConsole.runsource()`. A trailing newline is
needed for these expressions to be evaluated. I don't know why this is, the
docs don't mention anything.
From a python repl, the following samples show that a terminal newline allows
statements containing flow control to fully execute.
```
>>> import code
>>> repl = code.InteractiveConsole()
>>> repl.runsource("if True: print(1)")
True
>>> repl.runsource("if True: print(1)\n")
1
False
```
Notes:
From an interactive python repl, the output is not printed immediately. The
user is required to enter a blank line following the first.
```
>>> if True: print(1)
...
1
```
However, `python -c 'if True: print(1)'` works without needing a newline.
Reviewed By: JDevlieghere
Differential Revision: https://reviews.llvm.org/D127586
input_string = strip_and_check_exit(input_string)
repl = code.InteractiveConsole(local_dict)
if input_string:
+ # A newline is appended to support one-line statements containing
+ # control flow. For example "if True: print(1)" silently does
+ # nothing, but works with a newline: "if True: print(1)\n".
+ input_string += "\n"
repl.runsource(input_string)
elif g_run_one_line_str:
repl.runsource(g_run_one_line_str)
# RUN: %lldb -o 'script -lpython -- print("{}".format(1000+100+10+1))' 2>&1 | FileCheck %s
# RUN: %lldb -o 'script --language python -- print("{}".format(1000+100+10+1))' 2>&1 | FileCheck %s
# RUN: %lldb -o 'script --language=python -- print("{}".format(1000+100+10+1))' 2>&1 | FileCheck %s
+# RUN: %lldb -o 'script -lpython -- if True: print("{}".format(1000+100+10+1))' 2>&1 | FileCheck %s
# CHECK: 1111
# RUN: %lldb -o 'script --language invalid -- print("{}".format(1000+100+10+1))' 2>&1 | FileCheck %s --check-prefix INVALID