B )²ô`D>ã@sDdZddlmZddlmZddlmZddlZddlZddlZddl m Z ddl m Z ddl m Z dd l mZdd lmZdd lmZdd lmZe jZd ZdZdZdZdZdZdZdZdZdZ dZ!dZ"dZ#e!e"e#gZ$d3dd„Z%d4dd„Z&dd „Z'd!d"„Z(d5d#d$„Z)d%d&„Z*d6d(d)„Z+d*d+„Z,d,Z-d7d-d.„Z.d8d/d0„Z/d1d2„Z0dS)9z>Shared functions and classes for tfdbg command-line interface.é)Úabsolute_import)Údivision)Úprint_functionN)Úcommand_parser)Údebugger_cli_common)Ú tensor_format)Úcommon)Úops)Ú variables)ÚgfileiÐZblackZblueZcyanZgrayZgreenZmagentaZredZwhiteZyellowÚusÚmsÚsFcCsj|dkrt|ƒS|dkr"d|}n8|dkr8d|d}n"|dkrNd|d }n d |d }|rf|d 7}|S) aeGenerate a human-readable string representing number of bytes. The units B, kB, MB and GB are used. Args: num_bytes: (`int` or None) Number of bytes. include_b: (`bool`) Include the letter B at the end of the unit. Returns: (`str`) A string representing the number of bytes in a human-readable way, including a unit at the end. Niz%diz%.2fkg@i@z%.2fMg0Az%.2fGgÐAÚB)Ústr)Ú num_bytesZ include_bÚresult©rúU/home/dcms/DCMS/lib/python3.7/site-packages/tensorflow/python/debug/cli/cli_shared.pyÚbytes_to_readable_str7s  rcCs”|sdS|rJ|tkr td|ƒ‚t |¡}|}d |t dd|¡|¡Stttƒdtt  |d¡dƒƒ}t|}d |t dd|¡|¡Sd S) aWConvert time value to human-readable string. Args: value_us: time value in microseconds. force_time_unit: force the output to use the specified time unit. Must be in TIME_UNITS. Returns: Human-readable string representation of the time value. Raises: ValueError: if force_time_unit value is not in TIME_UNITS. Ú0zInvalid time unit: %sz {:.10g}{}g$@ééé z{:.3g}{}N) Ú TIME_UNITSÚ ValueErrorÚindexÚformatÚmathÚpowÚminÚlenÚintÚlog)Zvalue_usZforce_time_unitÚorderZ time_unitrrrÚtime_to_readable_strUs  "r%cs4d‰‡fdd„}|r,t |¡‰tj||dSdSdS)avProcess ranges highlight string. Args: ranges_string: (str) A string representing a numerical range of a list of numerical ranges. See the help info of the -r flag of the print_tensor command for more details. Returns: An instance of tensor_format.HighlightOptions, if range_string is a valid representation of a range or a list of ranges. Nc sBtj|jtd}x,ˆD]$\}}t |t ||k||k¡¡}qW|S)N)Zdtype)ÚnpÚzerosÚshapeÚboolÚ logical_orÚ logical_and)ÚxÚrZ range_startZ range_end)ÚrangesrrÚ ranges_filter€s z-parse_ranges_highlight..ranges_filter)Ú description)rZ parse_rangesrZHighlightOptions)Z ranges_stringr/r)r.rÚparse_ranges_highlightqs    r1cCs |rd|krd|diSiSdS)NÚcolsZ linewidthr)Z screen_inforrrÚ#numpy_printoptions_from_screen_infos  r3c CsÆ|rt ||¡}||} n|}|} d} |r–t |d¡} t | |¡WdQRXt d¡} | tj|dd7} | dtt  |¡j ƒ7} t  | t d¡g¡} |r¦|j |d<nt |d<tj|| d || ||d S) aGenerate formatted str to represent a tensor or its slices. Args: tensor: (numpy ndarray) The tensor value. tensor_name: (str) Name of the tensor, e.g., the tensor's debug watch key. np_printoptions: (dict) Numpy tensor formatting options. print_all: (bool) Whether the tensor is to be displayed in its entirety, instead of printing ellipses, even if its number of elements exceeds the default numpy display threshold. (Note: Even if this is set to true, the screen output can still be cut off by the UI frontend if it consist of more lines than the frontend can handle.) tensor_slicing: (str or None) Slicing of the tensor, e.g., "[:, 1]". If None, no slicing will be performed on the tensor. highlight_options: (tensor_format.HighlightOptions) options to highlight elements of the tensor. See the doc of tensor_format.format_tensor() for more details. include_numeric_summary: Whether a text summary of the numeric values (if applicable) will be included. write_path: A path to save the tensor value (after any slicing) to (optional). `numpy.save()` is used to save the value. Returns: An instance of `debugger_cli_common.RichTextLines` representing the (potentially sliced) tensor. NÚwbzSaved value to: Úbold)Ú font_attrz (%sB)ÚÚ thresholdT)Zinclude_metadataÚinclude_numeric_summaryÚauxiliary_messageÚnp_printoptionsÚhighlight_options)rZevaluate_tensor_slicer ZOpenr&ÚsaverÚRichLinerZStatÚlengthÚ#rich_text_lines_from_rich_line_listÚsizeÚ!DEFAULT_NDARRAY_DISPLAY_THRESHOLDrÚ format_tensor) ZtensorZ tensor_namer;Z print_allZtensor_slicingr<r9Z write_pathÚvalueZ sliced_namer:Z output_fileÚlinerrrrC–s2#    rCcCst td|tƒg¡S)zËGenerate a RichTextLines output for error. Args: msg: (str) The error message. Returns: (debugger_cli_common.RichTextLines) A representation of the error message for screen output. zERROR: )rr@ÚRLÚ COLOR_RED)ÚmsgrrrÚerrorÚs rIécCsNd|}|rt d|¡dg}nd}t|ƒt||ƒd|d|g}t |¡S)aÂGenerate a RichTextLines object that describes a recommended command. Args: command: (str) The command to recommend. description: (str) A description of what the command does. indent: (int) How many spaces to indent in the beginning. create_link: (bool) Whether a command link is to be applied to the command string. Returns: (RichTextLines) Formatted text (with font attributes) for recommending the command. ú r7r5ú:z )rÚMenuItemrFr@)Úcommandr0ÚindentÚ create_linkZ indent_strr6ÚlinesrrrÚ_recommend_commandésrRcCsdddddddg}t |¡S)z/Make an ASCII representation of the tfdbg logo.r7zTTTTTT FFFF DDD BBBB GGG z TT F D D B B G z TT FFF D D BBBB G GGz TT F D D B B G Gz TT F DDD BBBB GGG )rÚ RichTextLines)rQrrrÚget_tfdbg_logosrTz&======================================c Cs.t |¡}|st d¡g}nLg}xF|D]>}t |¡}t d¡} | t |t dd|¡¡7} | | ¡q&Wt |¡}t t ¡} |rŒ|  d¡nZ|  d|¡|  d¡|  d¡|   t d d „|Dƒ¡¡|  d¡|  d ¡|   |¡|  t ¡|  d¡|  d ¡|   t d ddd¡|   t dddd¡|   t ddƒ¡|   t ddƒ¡dg} |r g} xL|D]8} |  | ¡t dd| ¡}|  t dƒt | |ƒ¡qbWn |  d¡|   t | ¡¡|  d¡|   t dƒt dt dd¡ƒd¡|  d¡t ¡}| t d d ¡¡| t dd¡¡|| jtj<| S) aGGenerate formatted intro for run-start UI. Args: run_call_count: (int) Run call counter. fetches: Fetches of the `Session.run()` call. See doc of `Session.run()` for more details. feed_dict: Feeds to the `Session.run()` call. See doc of `Session.run()` for more details. tensor_filters: (dict) A dict from tensor-filter name to tensor-filter callable. is_callable_runner: (bool) whether a runner returned by Session.make_callable is being run. Returns: (RichTextLines) Formatted intro message about the `Session.run()` call. z (Empty)z Nzpf '%s'z4Running a runner returned by Session.make_callable()zSession.run() call #%d:r7z Fetch(es):cSsg|] }d|‘qS)z r)Ú.0rErrrú Gsz'get_run_start_intro..z Feed dict:z5Select one of the following commands to proceed ---->Úrunz1Execute the run() call with debug tensor-watchingT)rPzrun -nz4Execute the run() call without debug tensor-watchingz run -t zyExecute run() calls (T - 1) times without debugging, then execute run() once more with debugging and drop back to the CLIzrun -f zpKeep executing run() calls until a dumped tensor passes a given, registered filter (conditional breakpoint mode)z Registered filter(s):z run -f %sz * z (None)zFor more details, see zhelp.ÚhelpÚ.Úexit)rÚget_flattened_namesrr>Úget_graph_element_namerMÚappendr@rSÚ_HORIZONTAL_BARÚextendrRrFZappend_rich_lineZMenuÚ annotationsZ MAIN_MENU_KEY)Úrun_call_countÚfetchesÚ feed_dictZtensor_filtersÚis_callable_runnerZ fetch_linesZfeed_dict_linesZfeed_keyZ feed_key_nameZfeed_dict_lineÚoutÚ more_linesZ filter_namesZ filter_nameZcommand_menu_nodeÚmenurrrÚget_run_start_intros„                         rhcCsÐ|rdSd|}t|tjtjtjfƒr:|dt |¡7}n0tt  |¡ƒ}|dkr^|d|7}n |d|7}|sx|d7}nTt|ƒdkr¼xF|D],}|dt|t j ƒs¨t |d ƒs¬|n|j 7}qŠWn|d t|ƒ7}|S) aGet a short description of the run() call. Args: run_call_count: (int) Run call counter. fetches: Fetches of the `Session.run()` call. See doc of `Session.run()` for more details. feed_dict: Feeds to the `Session.run()` call. See doc of `Session.run()` for more details. is_callable_runner: (bool) whether a runner returned by Session.make_callable is being run. Returns: (str) A short description of the run() call, including information about the fetche(s) and feed(s). zrunner from make_callable()z run #%d: z1 fetch (%s); rz %d fetches; z %d fetch; z0 feedsz 1 feed (%s)Únamez%d feeds)Ú isinstancer ZTensorZ Operationr ÚVariablerr\r!r[ÚsixÚ string_typesÚhasattrri)rarbrcrdr0Z num_fetchesÚkeyrrrÚget_run_short_description‚s$    rpc Csît|dƒr t|jdƒr |jj}nd}dtddƒdg}t |¡}|dk rž| t dg¡¡| td |d d d ¡| td |dd d ¡| tddd d ¡n| t dg¡¡dd|dt t |ƒƒddt |ƒdddg }| t |¡¡|S)aGenerate formatted intro for TensorFlow run-time error. Args: tf_error: (errors.OpError) TensorFlow run-time error object. Returns: (RichTextLines) Formatted intro message about the run-time OpError, with sample commands for debugging. ÚopriNz&--------------------------------------z(!!! An error occurred during the run !!!Zblinkr7z,You may use the following commands to debug:zni -a -d -t %sz)Inspect information about the failing op.T)rPzli -r %sz+List inputs to the failing op, recursively.Últz6List all tensors dumped during the failing run() call.zCWARNING: Cannot determine the name of the op that caused the error.zOp name: %sz Error type: zDetails:) rnrqrirFrr@r_rSrRrÚtype)Ztf_errorZop_nameZ intro_linesrerfrrrÚget_error_intro²sH        rt)F)N)FNNFN)rJF)F)F)1Ú__doc__Ú __future__rrrrÚnumpyr&rlZtensorflow.python.debug.clirrrZtensorflow.python.debug.librZtensorflow.python.frameworkr Ztensorflow.python.opsr Ztensorflow.python.platformr r>rFrBZ COLOR_BLACKZ COLOR_BLUEZ COLOR_CYANZ COLOR_GRAYZ COLOR_GREENZ COLOR_MAGENTArGZ COLOR_WHITEZ COLOR_YELLOWZ TIME_UNIT_USZ TIME_UNIT_MSZ TIME_UNIT_Srrr%r1r3rCrIrRrTr^rhrprtrrrrÚsX               =  j -