|
|
|
|
|
\subsection{Mapfile structure}\label{subsec:mapfile-structure}
|
|
|
|
|
|
|
|
|
|
|
|
Mapfiles created by \texttt{ddrescue} use a compact text format to store meta information
|
|
|
|
|
|
about the data rescuing process and the information about the rescuing status of blocks\cite{ddrescue}.
|
|
|
|
|
|
A simple exemplary mapfile is shown below.
|
|
|
|
|
|
|
|
|
|
|
|
\begin{tcbcode}{Example mapfile}
|
|
|
|
|
|
# Mapfile. Created by GNU ddrescue version 1.27
|
|
|
|
|
|
# Command line: ddrescue /dev/sdb sdb.img sdb.map
|
|
|
|
|
|
# Start time: 2024-02-08 06:56:57
|
|
|
|
|
|
# Current time: 2024-02-08 06:56:59
|
|
|
|
|
|
# Copying non-tried blocks... Pass 1 (forwards)
|
|
|
|
|
|
# current_pos current_status current_pass
|
|
|
|
|
|
0x00970000 ? 1
|
|
|
|
|
|
# pos size status
|
|
|
|
|
|
0x00000000 0x02200000 +
|
|
|
|
|
|
0x02200000 0x00000001 -
|
|
|
|
|
|
0x02200001 0x3F690000 ?
|
|
|
|
|
|
\end{tcbcode}
|
|
|
|
|
|
|
|
|
|
|
|
The mapfile consists of three main parts:
|
|
|
|
|
|
|
|
|
|
|
|
The \textbf{heading comments} comments provide metadata about the mapfile, as the version of ddrescue or ddrescuelog that created it,
|
|
|
|
|
|
the command-line parameters used during the operation and the start time of the program.
|
|
|
|
|
|
If it was created by ddrescue, it also includes the current save time and a copy of the status message from the screen (e.g., copying, trimming, finished).
|
|
|
|
|
|
|
|
|
|
|
|
The first non-comment line, the \textbf{status line}, contains the position being tried in the input file,
|
|
|
|
|
|
a status character indicating the type of operation (e.g., copying, trimming, scraping)
|
|
|
|
|
|
and a positive integer denoting the current pass in the current phase.
|
|
|
|
|
|
The status line helps efficient resumption of copying or retrying phases.
|
|
|
|
|
|
|
|
|
|
|
|
The other non-comment lines each describe a \textbf{data block}, containing the starting position of the block in the input file,
|
|
|
|
|
|
the size of the block in bytes and a status character indicating the block’s state (e.g., copied, trimmed, scraped).
|
|
|
|
|
|
|
|
|
|
|
|
\subsection{Parsing in Python}\label{subsec:parsing-in-python}
|
|
|
|
|
|
The Python function \texttt{parse\_mapfile} implemented in \texttt{ddrescue\_tui\_parser.py} is designed to parse such mapfiles.
|
|
|
|
|
|
It takes as input the path to the mapfile and the desired resolution for the output array.
|
|
|
|
|
|
The function performs several steps of validation and parsing, using the module \texttt{numpy}\cite{numpy}:
|
|
|
|
|
|
|
|
|
|
|
|
\begin{itemize}
|
|
|
|
|
|
\item It checks the validity of the input parameters.
|
|
|
|
|
|
If the file path is not valid or the resolution parameters are not integers greater than 1, it returns with an error message.
|
|
|
|
|
|
\item It opens the mapfile and reads it line by line, distinguishing between comment lines and data lines.
|
|
|
|
|
|
The first line is used to check whether the opened file is actually a mapfile, otherwise the function returns an error message.
|
|
|
|
|
|
\item It parses the comment lines to extract general information about the process, such as the version of ddrescue that created the mapfile,
|
|
|
|
|
|
the command-line parameters used and the direction of the reading process (forwards or backwards).
|
|
|
|
|
|
\item It parses the data lines to create custom \texttt{MapEntry} objects, each representing a block of data.
|
|
|
|
|
|
The \texttt{MapEntry} object has a position, size, and status, which are extracted from the data line.
|
|
|
|
|
|
The status converted to a numerical value, which can later be used for plotting the information.
|
|
|
|
|
|
\item After parsing all lines of the mapfile, the function sorts the list of \texttt{MapEntry} objects by position
|
|
|
|
|
|
to retrieve information about the total size from the last block of data.
|
|
|
|
|
|
It then transforms this list into a 1D numpy array with the length of the desired plot resolution,
|
|
|
|
|
|
with each element representing the status of multiple disk blocks.
|
|
|
|
|
|
As the resolution of the plot is usually several orders of magnitude smaller than the number of blocks on the disk,
|
|
|
|
|
|
some information is lost when the status is compressed in this way.
|
|
|
|
|
|
The highest numerical status value determines which status a compressed block receives;
|
|
|
|
|
|
the values were assigned in such a way that the most interesting information is retained in the author's opinion.
|
|
|
|
|
|
\item Finally, the function reshapes the 1D numpy array into a 2D numpy array based on the provided resolution.
|
|
|
|
|
|
It also updates the current position in the array based on the current position given in the mapfile.
|
|
|
|
|
|
\end{itemize}
|
|
|
|
|
|
|
|
|
|
|
|
The function returns the 2D numpy array and a dictionary containing the general information.
|
|
|
|
|
|
This information will be used to visualize the status of different blocks of data in a disk imaging process
|
|
|
|
|
|
and provide context about the process, as described in the following sections.
|
