LESSON 01 / 13
Bash fundamentalsEvery command starts somewhere. Discover where you are in the filesystem.
THE CONCEPT
The shell always has a current folder. pwd prints its full path, so you know where relative paths will start.
pwdpwd → print working directory
YOUR UNIX TASK
Print the path of your current directory.
Hint 1. The command is an abbreviation for “print working directory”.
Hint 2. Type pwd and press Enter.
ONE WAY TO SOLVE IT
pwdType each command in the terminal to try it.
THE CONCEPT
cd changes the current directory. A relative path starts in your current folder; an absolute path starts at /. Use .. for the parent, ~ for home, and cd - to return to the previous directory.
cd projectscd → change directory · projects → folder to enter
YOUR UNIX TASK
Navigate into the website folder inside projects.
Hint 1. You can enter both folders in one path, separated by /.
Hint 2. From home, run cd projects/website.
ONE WAY TO SOLVE IT
cd projects/websiteType each command in the terminal to try it.
THE CONCEPT
cat writes a file’s contents to standard output: usually your terminal. It can also combine multiple files. Each lesson starts with a fresh practice filesystem in your home directory.
cat welcome.txtcat → print contents · welcome.txt → input file
YOUR UNIX TASK
Display the complete contents of welcome.txt.
Hint 1. Use cat followed by the file’s name.
Hint 2. Run cat welcome.txt from the home directory.
ONE WAY TO SOLVE IT
cat welcome.txtType each command in the terminal to try it.
THE CONCEPT
head prints the beginning of a file; tail prints the end. Both default to ten lines. Use -n followed by a number to choose how many lines to show.
tail -n 2 logs/app.logtail → last lines · -n 2 → two lines · logs/app.log → file
YOUR UNIX TASK
Print only the first two lines of welcome.txt using head.
Hint 1. You want the beginning, so choose head rather than tail.
Hint 2. Run head -n 2 welcome.txt.
ONE WAY TO SOLVE IT
head -n 2 welcome.txtType each command in the terminal to try it.
THE CONCEPT
grep keeps lines containing a pattern. By default, case matters: ERROR and error differ. Add -i to ignore case. This simulator teaches literal searches; full regular expressions are a later topic.
grep -i warn logs/app.loggrep → filter lines · -i → ignore case · warn → pattern
YOUR UNIX TASK
Display all three error lines from logs/app.log, including the lowercase one.
Hint 1. Search for error, and make the search case-insensitive.
Hint 2. Use grep -i error logs/app.log.
ONE WAY TO SOLVE IT
grep -i error logs/app.logType each command in the terminal to try it.
THE CONCEPT
A pipe, |, passes one command’s standard output to the next command’s standard input. wc -l counts newline characters. Combine it with a filter to count matching lines.
cat welcome.txt | wc -lcat → produce text · | → pass it along · wc -l → count lines
YOUR UNIX TASK
Use a pipeline to count all error lines in logs/app.log, ignoring case. The result should be 3.
Hint 1. Start with the case-insensitive error search from the previous lesson.
Hint 2. Pipe grep -i error logs/app.log into wc -l.
ONE WAY TO SOLVE IT
grep -i error logs/app.log | wc -lType each command in the terminal to try it.
THE CONCEPT
sort puts lines in order. uniq removes adjacent duplicates, so sort first when repeated values are scattered. sort -u is a useful shortcut. The practice shell uses a fixed character order; real sorting depends on locale.
sort data/names.txt | uniqsort → group matching lines · uniq → remove adjacent duplicates
YOUR UNIX TASK
Print the four unique names from data/names.txt in alphabetical order.
Hint 1. Group duplicate names by sorting before you remove them.
Hint 2. Use sort data/names.txt | uniq, or sort -u data/names.txt.
ONE WAY TO SOLVE IT
sort data/names.txt | uniqType each command in the terminal to try it.
THE CONCEPT
Plain sort compares characters, which can place 12 before 3. Add -n to compare numeric values. Add -r to reverse the ordering, and put the largest values first.
sort -n data/scores.txtsort → order lines · -n → compare as numbers
YOUR UNIX TASK
Print data/scores.txt from highest score to lowest.
Hint 1. You need both numeric sorting and reverse order.
Hint 2. Try sort -nr data/scores.txt.
ONE WAY TO SOLVE IT
sort -nr data/scores.txtType each command in the terminal to try it.
THE CONCEPT
The > operator writes standard output to a file, creating it if needed and replacing any existing contents. The terminal stays quiet because output went to the file. Use cat to inspect the result.
echo "My first note" > note.txtecho → produce text · > → write to a file · note.txt → destination
YOUR UNIX TASK
Write the text Hello terminal to a file named greeting.txt in your home directory.
Hint 1. echo prints text; > sends that text to a file.
Hint 2. Run echo "Hello terminal" > greeting.txt.
ONE WAY TO SOLVE IT
echo "Hello terminal" > greeting.txtType each command in the terminal to try it.
THE CONCEPT
While > replaces contents, >> adds output to the end of a file. Quotes keep multiple words together. Newlines come from the command producing the output, not from the redirection operator.
echo "Another line" >> note.txt>> → append output, preserving existing content
YOUR UNIX TASK
Create notes.txt with First step on the first line and Keep going on the second. Use >> to add the second line.
Hint 1. First use > to write First step, then >> to add Keep going.
Hint 2. Run echo "First step" > notes.txt, then echo "Keep going" >> notes.txt.
ONE WAY TO SOLVE IT
echo "First step" > notes.txtecho "Keep going" >> notes.txtType each command in the terminal to try it.
THE CONCEPT
find walks a directory tree. -type f selects files, and -name matches names. Quote a wildcard pattern so the real shell passes it to find instead of expanding it first.
find . -type f -name "*.txt". → current folder · -type f → files · -name → name pattern
YOUR UNIX TASK
Find README.md anywhere inside the projects directory.
Hint 1. Start the search at projects and use -name to select README.md.
Hint 2. Run find projects -name README.md.
ONE WAY TO SOLVE IT
find projects -name README.mdType each command in the terminal to try it.
THE CONCEPT
Filter the log, normalize its order, remove repeated lines, and save the result. Commands compose because they share a simple interface: text in, text out.
producer | filter | sort > report.txtThis is a pattern, not a runnable command. Choose the tools you have learned.
YOUR UNIX TASK
Save all error lines from logs/app.log, ignoring case, sorted with duplicates removed, to errors.txt in your home directory.
Hint 1. Begin with grep -i error logs/app.log.
Hint 2. Pipe the results to sort -u and redirect into errors.txt.
ONE WAY TO SOLVE IT
grep -i error logs/app.log | sort -u > errors.txtType each command in the terminal to try it.
THE CONCEPT
A second grep narrows the first result. grep -v excludes matching lines. Inspect the input before choosing filters; a match is text evidence, not proof of a root cause.
grep -i error logs/service.log | grep apiEach stage keeps only the lines that satisfy its filter.
YOUR UNIX TASK
Read logs/service.log. Save every API error, ignoring case, in original order to api-errors.txt. Save the number of those lines to error-count.txt (number only).
Hint 1. Build the filtered file first, then pipe its contents to wc -l.
ONE WAY TO SOLVE IT
grep -i error logs/service.log | grep api > api-errors.txtcat api-errors.txt | wc -l > error-count.txtType each command in the terminal to try it.
THE CONCEPT
uniq compares neighboring lines. sort brings equal IDs together; uniq -c counts each group and uniq -d keeps only repeated groups. Counting raw, unsorted input can miss repeats.
sort data/jobs.txt | uniq -cCounts describe occurrences, including the first copy.
YOUR UNIX TASK
Save only the repeated job IDs, once each and alphabetically, to repeated.txt. Save counts for every ID, alphabetically by ID, to frequencies.txt.
Hint 1. Both reports begin with sorted input. Consult man uniq for the two options.
ONE WAY TO SOLVE IT
sort data/jobs.txt | uniq -d > repeated.txtsort data/jobs.txt | uniq -c > frequencies.txtType each command in the terminal to try it.
THE CONCEPT
comm compares two sorted files. Its three columns contain lines unique to the first file, unique to the second, and shared by both. -23 suppresses columns two and three; -13 leaves only column two. Sort and deduplicate both inventories first.
comm -23 before-sorted.txt after-sorted.txtKeep entries that exist only in the first inventory.
YOUR UNIX TASK
Compare data/before.txt and data/after.txt. Save removed services to removed.txt and added services to added.txt, each alphabetically and without duplicates.
Hint 1. Create two sorted unique scratch files, then compare in both directions.
ONE WAY TO SOLVE IT
sort -u data/before.txt > before-sorted.txtsort -u data/after.txt > after-sorted.txtcomm -23 before-sorted.txt after-sorted.txt > removed.txtcomm -13 before-sorted.txt after-sorted.txt > added.txtType each command in the terminal to try it.
THE CONCEPT
mkdir creates a destination folder. cp copies a file; mv relocates or renames it. Quote names containing spaces. This simulator supports one file at a time and rejects existing destinations to prevent accidental overwrites; real cp and mv can overwrite files.
cp "inbox/project brief.txt" delivery/brief.txtCopy the source to a deliberately named destination.
YOUR UNIX TASK
Create delivery. Copy inbox/project brief.txt to delivery/brief.txt, retaining the original. Move inbox/results.txt to delivery/results.txt. Leave inbox/scratch.txt untouched. Save a sorted list of file paths under delivery to manifest.txt.
Hint 1. Create the folder before copying or moving. Use find to build the manifest.
ONE WAY TO SOLVE IT
mkdir deliverycp "inbox/project brief.txt" delivery/brief.txtmv inbox/results.txt delivery/results.txtfind delivery -type f | sort > manifest.txtType each command in the terminal to try it.
THE CONCEPT
cut -d chooses a one-character delimiter; -f selects a field, numbered from one. This exercise uses simple pipe-delimited records, not quoted CSV. Extract service names, count them, then sort counts numerically in descending order.
cut -d "|" -f 2 data/requests.txtEach row is request ID | service | outcome, with no header.
YOUR UNIX TASK
Save counts per service to service-report.txt, highest count first. Include successful and failed requests. Preserve the input file.
Hint 1. Extract field two, sort it, count groups, then sort the counts numerically.
ONE WAY TO SOLVE IT
cut -d "|" -f 2 data/requests.txt | sort | uniq -c | sort -nr > service-report.txtType each command in the terminal to try it.
THE CONCEPT
The logs use comparable, zero-padded timestamps and request IDs. A timeline requires matching the same request across files and ordering all of its events. A nearby failure from another request may be unrelated. Read the evidence before naming the cause.
cat logs/gateway.log logs/database.logCombine sources before filtering and ordering.
YOUR UNIX TASK
Investigate request req-42. Save all its events from both logs chronologically to timeline.txt. Then save just its ERROR events to incident-errors.txt. Write the failing dependency name (database) to dependency.txt after reviewing the timeline.
Hint 1. The request ID connects the files. Keep context events in the timeline, then derive the error-only report.
ONE WAY TO SOLVE IT
cat logs/gateway.log logs/database.log | grep req-42 | sort > timeline.txtgrep ERROR timeline.txt > incident-errors.txtecho database > dependency.txtType each command in the terminal to try it.
THE CONCEPT
Read data/rules.txt first. These fixtures use a deliberately narrow rule: rows containing |active| are eligible unless the address is MISSING. Field three holds the address. This is a structured exercise, not a general email validator. Keep rejected records instead of silently dropping them. Use grep -F for literal punctuation such as the pipe delimiter.
cat data/rules.txtDiscover the acceptance rule before writing the pipeline.
YOUR UNIX TASK
Save unique eligible addresses alphabetically to recipients.txt. Save all rejected rows to rejected.txt in alphabetical order. Keep data/members.txt unchanged.
Hint 1. Separate inactive rows from active rows with missing addresses; combine those rejection groups before sorting.
ONE WAY TO SOLVE IT
grep -F "|active|" data/members.txt | grep -v MISSING | cut -d "|" -f 3 | sort -u > recipients.txtgrep -Fv "|active|" data/members.txt > rejected-work.txtgrep -F "|active|" data/members.txt | grep MISSING >> rejected-work.txtsort rejected-work.txt > rejected.txtType each command in the terminal to try it.
THE CONCEPT
A shell script is a text file of commands. Single quotes around an echo argument store pipes and redirection literally. sh report.sh reads and runs those commands. Our bounded runner supports up to 20 simple command lines, comments, and pipelines; variables, arguments, loops, and nested scripts are outside this exercise. It stops on errors, unlike a default real sh script.
echo 'cat logs/service.log | wc -l > total.txt' > report.shThe outer > writes the script. The quoted > is executed later by sh.
YOUR UNIX TASK
Create report.sh in your home folder. It must generate script-errors.txt from all error lines in logs/service.log, ignoring case, sorted and deduplicated. Run sh report.sh. Your script must generate a fresh report when the log contents change.
Hint 1. Store the working filter-and-sort pipeline inside single quotes, then execute the saved file.
ONE WAY TO SOLVE IT
echo 'grep -i error logs/service.log | sort -u > script-errors.txt' > report.shsh report.shType each command in the terminal to try it.
THE CONCEPT
A pipeline can run successfully and still answer the wrong question. Read the supplied broken command. Check which lines survive filtering, whether repeats are adjacent when counted, and whether the final sort uses numbers in the intended direction.
cat broken-pipeline.txtRead the attempted workflow, then inspect stages in isolation.
YOUR UNIX TASK
Repair the pipeline described in broken-pipeline.txt. Save the two most frequent error messages and their counts to top-errors.txt, highest count first. Match ERROR and error. Count the message field (field three), not the whole record.
Hint 1. There are three logic mistakes: case-sensitive filtering, counting before grouping, and ascending numeric sorting.
ONE WAY TO SOLVE IT
grep -i error logs/events.log | cut -d "|" -f 3 | sort | uniq -c | sort -nr | head -n 2 > top-errors.txtType each command in the terminal to try it.
Try the task in the terminal. Mistakes are part of learning.
Each lesson starts fresh. These files exist only in this browser session.
A focused simulator. No commands run on your computer.
Your practice shell is ready. Type a Unix command below. Use help to see what is available.
LEARN IN PARALLEL
Same idea, different shells. These examples assume equivalent practice files; you don’t need to enter them.
cdWith no argument, CMD’s cd shows the current drive and directory. In Bash, cd with no argument goes home instead.
Microsoft reference ↗Get-LocationShows the current PowerShell location. On the filesystem, the Path property is the directory path.
Microsoft reference ↗cd /d projects\websiteChanges directory; /d also changes the active drive when a drive is included. Use cd .. to go up and cd /d "%USERPROFILE%" to go home.
Microsoft reference ↗Set-Location .\projects\websiteChanges the current location. Set-Location .. goes up; Set-Location ~ goes home.
Microsoft reference ↗type welcome.txtPrints the text file. Enclose a filename in double quotes when it contains spaces.
Microsoft reference ↗Get-Content .\welcome.txtReads a text file as a sequence of lines. These strings can flow into another cmdlet.
Microsoft reference ↗more welcome.txtRelated tool: displays a file one screen at a time. CMD has no simple built-in head equivalent that selects exactly the first two lines; use the PowerShell example.
Microsoft reference ↗Get-Content .\welcome.txt -TotalCount 2Selects the first two lines. Use -Tail 2 instead for the last two lines.
Microsoft reference ↗findstr /i /l /c:"error" logs\app.log/i ignores case; /l uses literal matching; /c: keeps the search text together. CMD’s findstr options use slashes.
Microsoft reference ↗Select-String -Path .\logs\app.log -Pattern 'error' -SimpleMatchCase-insensitive by default. -SimpleMatch treats the pattern as literal text. Results are MatchInfo objects containing the matching line, file, and line number.
Microsoft reference ↗type logs\app.log | find /i /c "error"The pipe sends text to find. /i ignores case and /c counts matching lines. Here find searches text, unlike Unix find, which searches paths.
Microsoft reference ↗Select-String -Path .\logs\app.log -Pattern 'error' -SimpleMatch | Measure-Object | Select-Object -ExpandProperty CountMeasure-Object counts the matching objects. Select-Object extracts the count. PowerShell cmdlets pass objects through the pipeline, not just text.
Microsoft reference ↗sort /unique data\names.txtOn Windows versions supporting /unique, sort can remove duplicates. Check sort /? on the target machine; PowerShell provides the more portable Windows approach.
Microsoft reference ↗Get-Content .\data\names.txt | Sort-Object -UniqueSorts the lines and removes duplicates. String sorting is case-insensitive by default and may differ from Unix locale-based sorting.
Microsoft reference ↗sort /r data\scores.txtRelated tool: /r reverses text order, but CMD sort has no numeric-sort option. It does not solve this numeric task. Use PowerShell to compare values as numbers.
Microsoft reference ↗Get-Content .\data\scores.txt | Sort-Object { [int]$_ } -DescendingCasts each line to an integer for sorting, then puts the largest first. $_ is the current pipeline item; every line here must be a valid integer.
Microsoft reference ↗>greeting.txt echo Hello terminalWrites the line to a file and replaces existing contents. Redirection is placed first to avoid an accidental trailing space in the message. CMD echo would print surrounding quotation marks.
Microsoft reference ↗Set-Content -Path .\greeting.txt -Value 'Hello terminal' -Encoding utf8Creates or replaces the file. Set-Content makes the intent explicit. UTF-8 BOM behavior differs between Windows PowerShell 5.1 and PowerShell 7.
Microsoft reference ↗>notes.txt echo First step
>>notes.txt echo Keep goingRun these as two separate lines. > replaces the file; >> appends, keeping the first line.
Microsoft reference ↗Set-Content .\notes.txt 'First step' -Encoding utf8
Add-Content .\notes.txt 'Keep going' -Encoding utf8Set-Content starts the file; Add-Content appends the next line. These examples specify the same encoding for both writes.
Microsoft reference ↗dir projects\README.md /s /b /a:-d/s searches subdirectories, /b gives bare paths, and /a:-d selects files. CMD find is a text-search command, not the Unix file finder.
Microsoft reference ↗Get-ChildItem .\projects -Recurse -Filter README.md -FileRecursively finds matching files. Results are file objects; add | Select-Object -ExpandProperty FullName if you want only path strings.
Microsoft reference ↗findstr /i /l /c:"error" logs\app.log | sort /unique > errors.txtFilters, sorts, removes duplicates, and saves text. Requires a sort version supporting /unique. Windows sort ordering and case handling can differ from the Unix exercise.
Microsoft reference ↗Select-String -Path .\logs\app.log -Pattern 'error' -SimpleMatch | ForEach-Object { $_.Line } | Sort-Object -Unique -CaseSensitive | Set-Content .\errors.txt -Encoding utf8Extracts the Line property before sorting and writing. -CaseSensitive preserves differences between differently cased lines. Culture-based ordering can still differ from Unix.
Microsoft reference ↗findstr /i /l /c:"error" logs\service.logRelated technique, not a complete solution to this project. /i ignores case; /l uses literal matching; /c: keeps the search text together. CMD’s findstr options use slashes.
Microsoft reference ↗Select-String -Path .\logs\service.log -Pattern 'error' -SimpleMatchRelated technique, not a complete solution to this project. Case-insensitive by default. -SimpleMatch treats the pattern as literal text. Results are MatchInfo objects containing the matching line, file, and line number.
Microsoft reference ↗sort /unique data\jobs.txtRelated first step: produce a unique job inventory. This does not select only repeated jobs or count occurrences. On Windows versions supporting /unique, sort can remove duplicates. Check sort /? on the target machine; PowerShell provides the more portable Windows approach.
Microsoft reference ↗Get-Content .\data\jobs.txt | Group-Object -NoElement | Sort-Object NameGroups matching IDs and reports their counts, even when repeats are not adjacent. Unlike uniq, input need not be sorted first. This is the frequency report step.
Microsoft reference ↗sort /unique data\before.txtPreparation step: sort and deduplicate the first release inventory before comparing. On Windows versions supporting /unique, sort can remove duplicates. Check sort /? on the target machine; PowerShell provides the more portable Windows approach.
Microsoft reference ↗$before = Get-Content .\data\before.txt | Sort-Object -Unique
$after = Get-Content .\data\after.txt | Sort-Object -Unique
Compare-Object -ReferenceObject $before -DifferenceObject $after<= marks entries only in the reference (removed); => marks entries only in the difference set (added). Outputs objects rather than comm’s text columns.
Microsoft reference ↗move inbox\results.txt delivery\results.txtThe delivery directory must already exist. Moves the results file; the copy and manifest steps are separate. CMD can prompt before overwriting an existing target.
Microsoft reference ↗Move-Item -Path .\inbox\results.txt -Destination .\delivery\results.txtThe delivery directory must already exist. Relocates the file; this covers the move step only. The source no longer remains in inbox.
Microsoft reference ↗type data\requests.txtInspection step only: read the source before constructing a report. The PowerShell example demonstrates grouping and numeric counts; CMD type does not aggregate records.
Microsoft reference ↗Get-Content .\data\requests.txt | ForEach-Object { $_.Split('|')[1] } | Group-Object -NoElement | Sort-Object Count -DescendingExtracts the service with a zero-based array index, groups it, and sorts numeric Count properties. Displays a table; saving a report is a separate step.
Microsoft reference ↗findstr /i /l /c:"req-42" logs\gateway.logRelated technique, not a complete solution to this project. /i ignores case; /l uses literal matching; /c: keeps the search text together. CMD’s findstr options use slashes.
Microsoft reference ↗Select-String -Path .\logs\gateway.log -Pattern 'req-42' -SimpleMatchRelated technique, not a complete solution to this project. Case-insensitive by default. -SimpleMatch treats the pattern as literal text. Results are MatchInfo objects containing the matching line, file, and line number.
Microsoft reference ↗findstr /i /l /c:"|active|" data\members.txtEligibility filter only; missing addresses still need to be excluded. /i ignores case; /l uses literal matching; /c: keeps the search text together. CMD’s findstr options use slashes.
Microsoft reference ↗Select-String -Path .\data\members.txt -Pattern '|active|' -SimpleMatchEligibility filter only; missing addresses still need to be excluded. Case-insensitive by default. -SimpleMatch treats the pattern as literal text. Results are MatchInfo objects containing the matching line, file, and line number.
Microsoft reference ↗type report.shInspect the saved Unix script as text. This does not execute it. CMD batch files and PowerShell scripts use different syntax; this Unix script needs a compatible Unix shell.
Microsoft reference ↗Get-Content .\report.shInspect the saved Unix script as text. This does not execute it. CMD batch files and PowerShell scripts use different syntax; this Unix script needs a compatible Unix shell.
Microsoft reference ↗type logs\events.logInspection step only: read the source before constructing a report. The PowerShell example demonstrates grouping and numeric counts; CMD type does not aggregate records.
Microsoft reference ↗Select-String -Path .\logs\events.log -Pattern 'error' -SimpleMatch | ForEach-Object { $_.Line.Split('|')[2] } | Group-Object -NoElement | Sort-Object Count -Descending | Select-Object -First 2Select-String ignores case by default. Extracts the message, groups occurrences, then selects the two highest counts. Displays objects; this does not write top-errors.txt.
Microsoft reference ↗